Tesla Regen, Brakes and Sudden Acceleration

by
Ronald A. Belt
Plymouth, MN 55447
1 June 2020

Abstract: EDR log data from a Tesla sudden acceleration incident is presented. To explain the
EDR data, the operation of Tesla’s drive motor control system and braking system are examined.
As expected, friction braking and regen are completely separate with no blending. The braking
system, however, includes several vehicle stability control functions that have a profound effect
on regen operation in the presence of wheel slip, such as stopping regen when going over bumps
and while turning corners. One of these slip control functions can cause the drive motor to
speed up if regen is causing slip in the rear drive wheels that can lead to an oversteer or
understeer. This same slip control function can be misled by a defective brake light switch to
confuse a brake-induced deceleration for a regen-induced deceleration, in which case the harder
a driver presses on the brake pedal, the larger a positive motor torque is produced. This is
believed to be the cause of sudden acceleration in over 70% of Tesla vehicles.
I. Introduction
EDR data from a Tesla M3 sudden acceleration incident reveals several inconsistencies between
the EDR data, the driver’s testimony, and Tesla’s own analysis of the log data of the incident. To
explain these inconsistencies, the designs of Tesla’s drive motor control system and braking
system were examined. The resulting explanation revealed that the cause of the sudden
acceleration lies in the vehicle’s braking system and how it interacts with the regen system.
II. EDR data
The EDR data for this incident will now be described. The incident involved a driver entering
the driveway of her home with the intent of parking in the attached garage. As the vehicle
neared the end of a 90° right-hand turn into the driveway, the driver had her right foot hovering
over the brake pedal in preparation for the garage door to fully open. Suddenly, the vehicle’s
drive motor revved up, causing the vehicle to leap forward and veer to the left. Simultaneously,
she pressed the brake pedal, but was unable to stop the vehicle in time before it struck a brick
wall between the two garage doors, causing minor damage to the right front corner of her
vehicle. The vehicle was a 2019 Tesla M3 with a single rear drive motor. The vehicle was in the
HOLD mode with STANDARD regen throughout the incident. (Figure 1).

Fig 1. Model 3 control panel showing options Fig 2. Vehicle path for the incident
for driving mode and regen braking as inferred from the EDR data

Tesla Regen, Brakes and 1 R. Belt

Sudden Acceleration 1 June 2020
Figure 2 shows the vehicle path during the incident. The street and driveway were both flat and
on the same horizontal level. Figure 3 shows the EDR data for the pre-crash values of the
accelerator pedal (%), rear motor speed (RPM), and vehicle speed (MPH). The accelerator pedal
data shows that a value of 80% was reached one second prior to the crash. The vehicle speed
data shows that the speed changed from 6 MPH before the crash to 14 MPH at the time of the
crash. These alone might suggest that the driver pressed on the accelerator pedal to cause the
crash. However, the pre-crash accelerometer data tell a different story.

Figure 3. EDR data showing pre-crash values of accelerator

pedal (%), rear motor speed (RPM), and vehicle speed (MPH)

Figure 4 shows the pre-crash accelerometer data. The longitudinal accelerometer data show
that the vehicle had a rapid deceleration of -0.5 g’s from +0.1 g to -0.4 g one second before the
crash. Since the Model 3’s regeneration is limited to -0.2 g’s (increased to -0.3 g’s by software
update 2018.42 v9 on October 25, 2018), this higher -0.5 g deceleration could not have been
caused by the vehicle’s regen system. Instead, it can only have been caused by the vehicle’s
braking system. This conclusion is supported by the high resolution log data described by Tesla
which shows that the brakes were applied even though the EDR data says they were not (Fig 5).
Therefore, the driver was pressing on the brake pedal at the time the drive motor revved up,
exactly as she described.
Comparison of the accelerator pedal data to the longitudinal acceleration data shows that the
increase and decrease in the accelerator pedal data are coincident with the decrease and
increase in the longitudinal acceleration data, which is determined by the depression of the
brake pedal. This suggests the possibility that the accelerator pedal may have been pressed at

Tesla Regen, Brakes and 2 R. Belt

Sudden Acceleration 1 June 2020
the same time as the brake pedal. However, Tesla has often stated in public that if this happens
the brake pedal will always win, implying that a brake pedal override function exists in all Tesla
vehicles that will cancel the accelerator command.1, 2 The brake pedal is also higher than the
accelerator pedal, preventing the accelerator pedal from being pressed to 88% while also
pressing the brake pedal with the same foot. Therefore, the accelerator pedal in this case could
not have been pressed at the same time as the brake pedal. So, what could have caused the drive
motor to rev up if the driver was pressing on the brake pedal and not on the accelerator pedal?

Figure 4. Pre-crash accelerometer data. An ISO 8855 coordinate system is used

in which the longitudinal x-axis points forward, the y-axis points to the driver’s left,
and the z-axis points up. The steering wheel angle is positive when turned to the right.
The longitudinal acceleration data in Figure 4 also shows no evidence that regen is occurring as
the vehicle is turning into the driveway before the drive motor revs up. This observation is
supported by the vehicle speed data of Figure 3, which shows that the vehicle speed remains
constant at 6 MPH before the drive motor revs up. Yet, the driver maintains that the vehicle was
in the HOLD mode with STANDARD regen throughout the incident. If the driver is correct,
then why isn’t regen apparent in the data of these two sensors?

Tesla Regen, Brakes and 3 R. Belt

Sudden Acceleration 1 June 2020
More inconsistencies become apparent if we compare the accelerometer data while aligned in
time, as shown in Figure 6. Both the steering wheel data and the yaw rate data agree with each
other and show that the vehicle was turning to the right into the driveway. At about 76° into the
90° turn, as the drive motor revs up to cause the vehicle to accelerator forward, both the steering
wheel data and the yaw rate data change direction, indicating a veer to the left. The acceleration
in the forward direction appears to be causing the veer to the left. This change of direction in
the yaw data reconciles the difference between the driver’s testimony and the EDR data and
Tesla’s letter as shown in table of Figure 5.

Figure 5. Comparison of the driver’s Figure 6. Accelerometer plots aligned in time

testimony with the EDR data and a Tesla show that the peaks roughly align at -1.0 seconds
letter describing the high resolution log data.
A closer look at the steering wheel angle data and the yaw rate data reveals that the yaw rate is
still increasing while the steering wheel angle is leveling off or even decreasing. This suggests
that front wheel slip is occurring to cause an oversteer condition, which could activate the

Tesla Regen, Brakes and 4 R. Belt

Sudden Acceleration 1 June 2020
vehicle’s stability control system. The normal response of a vehicle stability control system to an
oversteer condition is to brake the outside front wheel to cause a yaw moment in the direction
opposite to the oversteer direction. In this case, during an oversteer to the right, the left front
wheel would be braked, causing the vehicle to yaw to the left. This is exactly what is observed.
This appears to explain the driver’s testimony that the vehicle “turned to the left” to cause
damage to the right front corner of her vehicle.
The lateral acceleration data also show an increase and decrease in the lateral acceleration that
corresponds to the increase and decrease in the longitudinal acceleration, as one would expect
when accelerating forward while turning to the right. The roll data also shows a roll to the left
and then back to the right in agreement with the lateral acceleration data. The EDR data taken
together show that a sequence of events occurs that begins with a 90° turn to the right with no
regen occurring, followed by the brakes being applied, during which the drive motor speeds up
to cause a forward acceleration that produces an oversteer situation that activates the electronic
stabilization system, which brakes the front left wheel to cause the vehicle to veer to the left.
It is interesting that the vehicle’s longitudinal (forward) acceleration continues to increase until
the vehicle crashes into the garage, even though the accelerator pedal reading goes to zero over a
half second before the crash. One would expect that the vehicle’s velocity would remain
constant or even decrease after the accelerator pedal reading goes to zero, but the longitudinal
acceleration should cease, or even go negative with regen, when the accelerator pedal is
released. Also, the rear motor speed data show that the drive motor speed remains high even
after the accelerator pedal is released. This suggests that the drive motor is still producing
torque to accelerate the vehicle in the forward direction at the time of the crash, which is over a
half second after the accelerator pedal is released. How can this happen?
Finally, the EDR data show that the brakes were not applied during the incident while the driver
maintains that they were. However, the longitudinal accelerometer data and Tesla’s letter based
on the high resolution log data verify that the brakes were applied in support of the driver’s
testimony. Likewise, the EDR data show that the ABS system did not engage while both the
accelerometer data and Tesla’s letter based on the high resolution log data verifies that the ABS
system did engage.
In summary, the EDR data taken together show that a sequence of events takes place that begins
with a 90° turn to the right with no regen occurring, followed by the brakes being applied,
during which the drive motor speeds up to cause a forward acceleration that produces an
oversteer situation that activates the electronic stabilization system, which brakes the front left
wheel to cause the vehicle to veer to the left. However, we are left with a number of questions as
to how this could happen:
1. Why does regen appear to be absent in the vehicle velocity data and the longitudinal
acceleration data during the turn when the driver claims that the vehicle was in the
HOLD mode at the time and the driver was not pressing on the accelerator pedal?
2. Why does the single rear drive motor speed increase at exactly the same time as the
brakes are being applied?
3. If the driver had her foot on the brake pedal to cause the -0.5 g longitudinal acceleration,
then how could she also have pressed on the accelerator pedal at the same time to cause
the drive motor to speed up? Tesla has stated many times in writing that when the
accelerator pedal and the brake pedal are pressed at the same time, the brake pedal
always wins. Therefore, the speeding up of the drive motor cannot be due to the driver
pressing on the accelerator pedal, and must have been caused by the vehicle itself.
4. Why does the drive motor speed remain high and the vehicle continue to accelerate
forward even after the accelerator pedal reading has decreased back to zero?

Tesla Regen, Brakes and 5 R. Belt

Sudden Acceleration 1 June 2020
 5. Why did the vehicle steer to the left as the motor speed increases, even though the driver
was making a right hand turn?
6. Why does the EDR data show that the brake pedal was not pressed even though the
driver maintains that she did press the brake pedal and even though Tesla’s letter based
on the high resolution log data agrees with the driver?
7. If the driver never pressed the accelerator pedal, then why does the EDR data show that
the accelerator pedal was pressed?
8. Why does the EDR data show that the ABS system did not engage when both the
accelerometer data and Tesla’s letter based on the high resolution log data verify that the
ABS system did engage?
To seek answers to these questions, the designs of Tesla’s drive motor control system and
braking system were examined.
III. Tesla Drive Motor Control System
For the benefit of readers not familiar with Tesla’s one pedal driving (OPD) system, we will start
by describing how this system operates. This system allows control over both acceleration and
regen braking by varying the depression of the accelerator pedal. During regen braking,
deceleration of the vehicle causes the vehicle’s drive motor to act as a generator that recharges
the drive battery. This recharging is essential to obtaining the high number of miles on a single
battery charge that Tesla claims and drivers require. If regen is controlled only by the brake
pedal, then such high mileage would not be possible because infrequent braking events would
not charge the battery often enough unless the driver rides the brake pedal, which is tiring to the
driver and hazardous to other vehicles.
Figure 7 shows how Tesla’s one pedal driving system operates. When the accelerator pedal is
pressed to the floor the driver gets 100% acceleration. As the accelerator pedal is eased up, the
driver gets less acceleration. As the accelerator pedal is eased up further, eventually the

Figure 7. Tesla’s one pedal driving system uses the

accelerator pedal for both accelerating and braking. 3
vehicle acceleration ceases and turns into a mild deceleration with a small amount of battery
charging. This deceleration with battery charging is called regen. It feels to the driver like a
combustion engine vehicle with a stick shift while decelerating in second gear. In between the

Tesla Regen, Brakes and 6 R. Belt

Sudden Acceleration 1 June 2020
acceleration and deceleration points is a “sweet spot” that allows vehicle coasting. When the
accelerator pedal is fully released, the driver gets the maximum deceleration with a maximum
amount of battery charging, or maximum regen. The deceleration value with maximum regen is
about 0.2 g’s on older Tesla’s and 0.3 g’s on newer Tesla’s. With 0.3 g’s the deceleration level
feels like a combustion engine vehicle with a stick shift while decelerating in first gear. The
feature of one pedal driving that all Tesla drivers appreciate is that torque can be modulated
continuously between maximum positive torque and negative braking torque while keeping
one’s foot on the accelerator pedal. The only time that the driver needs to press the brake pedal
is when he wants to come to a complete stop or when he needs to do emergency braking at a
level higher than 0.3 g’s.
Figure 8 shows the first step in Tesla’s drive motor control system. In this step a pedal map
translates the accelerator pedal position set by the driver into a requested motor torque value.
The pedal map consists of a two-dimensional look-up table accessed by the accelerator pedal
position signal. Torque values for vehicle velocities in between velocities on the horizontal axis
are obtained by interpolating between torque values for vehicle velocities that are in the map.
The accelerator pedal position signal is an 8-bit sign-magnitude integer obtained by calculation
from the dual accelerator pedal position sensors. The torque values are 8-bit sign-magnitude
integers in the Tesla Model 3 and 16-bit sign-magnitude integers in the Tesla Model S and X.
Torque values are produced every 100 milliseconds. Equations for deriving the mapping values
can be found in References 2 and 3.4,5

Figure 8. A pedal map translates the accelerator pedal position into a requested
torque value. This typical map from a non-Tesla vehicle shows deceleration values
in g’s produced by the requested torques when acting on a constant vehicle inertia.4
Figure 9 shows the second step in Tesla’s drive motor control system. In this step a drive motor
torque map translates the requested torque values from the pedal map into torque and magnetic
field flux commands to the drive motor. The map consists of a two-dimensional look-up table
Tesla Regen, Brakes and 7 R. Belt
Sudden Acceleration 1 June 2020
accessed by the pedal map torque and the vehicle velocity. Table outputs for pedal map torques
and vehicle velocities lying in between those on the two axes are obtained by two-dimensional
interpolation of values that are in the map. Figure 9 shows only the resulting torque commands.
Positive motor torque commands in the forward driving quadrant enable driving in the forward
direction while drawing power from the battery. They may vary in magnitude from o g’s to 1.0
g’s at 100% of motor torque. The dashed line in Figure 3 shows a typical acceleration profile,
which can be any compound curve in the forward driving quadrant. Negative motor torque
commands in the forward braking quadrant enable regen braking with associated recharging of
the drive battery. They are limited by several factors to only a small part of the available motor
torque profile, as discussed below. The remainder of the forward braking quadrant goes unused.
The dashed line in Figure 3 shows a typical regen path in the forward braking quadrant.

Figure 9. A drive motor torque map translates the requested torque values from
the pedal map into torque and magnetic field flux commands to the drive motor.
Figure 10 shows in greater detail the red regen portion of the motor torque map of Figure 9. The
maximum braking torque is limited by the maximum deceleration level of -0.2 g’s to -0.3 g’s by
the desire to avoid vehicle instability when suddenly applying regen on roads with degraded
friction coefficients. At this maximum negative torque, as the motor speed increases, its back
emf also increases until it becomes equal to the excitation voltage, at which point the motor
speed cannot be increased. To increase the motor speed further, the emf is reduced by reducing
the motor’s field current in proportion to the inverse speed . Since torque is proportional to
motor current, the torque decreases as 1/ in this field weakening region. The power is
proportional to torque times the vehicle speed, and remains constant in this field weakening
region.

Tesla Regen, Brakes and 8 R. Belt

Sudden Acceleration 1 June 2020
 Figure 10. Detailed view of the regen portion of the motor torque map
showing the various factors limiting regen6
At any given torque below the maximum braking torque, as the motor speed decreases the regen
power and braking torque decrease in proportion to the motor speed . The rate of decrease is
limited by the maximum regen power, which must be kept less than some maximum value
determined by battery charging circuitry, as shnown by the dotted line in Figure 10. If this
maximum power is increased by a different design, one eventually becomes limited by the back
emf voltage which must remain above some minimum value for effective batttery charging, as
shown by the line with triangles in Figure 10. Below this minimum emf voltage, braking toque is
possible only by drawing power from the battery, which is known as plug braking in the
technical literature. It is important to note that the motor will still operate in all areas of the
forward braking quadrant (second quadrant). Operation in the desired regen braking region
(red area in Figure 10) is obtained only by limiting the torque request at a given speed.
Within the regen braking region as described by the above limits, it is possible to obtain any
desired braking torque and regen power by providing the drive motor with the appropriate
torque and field flux commands. Vehicle deceleration follows a path anywhere in this region as
the vehicle speed, which is directly proportional to the motor speed, decreases. At vehicle
speeds below 5 mph, induction motors cannot generate any effective braking torque or regen
power because their rotor fields are limited by induction currents which become too low for
practical use. However, motors having rotors with embedded permanent magnets can still
create rotor fields large enough to allow operation below 5 mph. Tesla introduced such internal
permanent magnet (IPM) motors in their Model 3 vehicles, and plans to use them in all new
versions of their Model S and Model X lines.
Figure 11 shows the operating points used in a drive motor torque map while following a typical
urban driving cycle. The driving cycle is EPA’s standard Urban Dynamometer Driving Schedule
(UDDS) driving cycle as shown in Figure 12. The operating points are shown in red, and clearly
show the positive acceleration torques and negative braking torques used. The linear variation
of braking torque with vehicle speed at low vehicle speeds is readily apparent.

Tesla Regen, Brakes and 9 R. Belt

Sudden Acceleration 1 June 2020
 Figure 11. Efficiency contours and operating Figure 12. Vehicle speed profile used in
points in a drive motor torque map while EPA’s Urban Dynamometer Driving
following EPA’s UDDS driving cycle.7 Schedule (UDDS) cycle.8
It is well known that many Tesla vehicles have dual drive motors. Figure 13 shows how regen
behaves in a Tesla Model 3 with dual drive motors as it slows down from 100 kph to 0 kph. As
the vehicle speed decreases, all the regen torque and power are provided by the rear drive motor
only, which produces the maximum regen possible. The braking torque corresponds to -0.3 g’s,
the maximum for a Model 3 with software update 2018.42 v9. At low speeds the regen torque
and power decrease linearly with speed as expected. Figure 14 shows the same situation as
Figure 13 except when there is snow on the road that creates slip between the road and the tires.
When slip is present, one finds that some of the rear motor braking torque is transferred to the
front motor, with the sum of the two torques remaining constant near the original maximum
value. Why is this happening?

Figure 13. Regen behavior of a Tesla Model 3 Figure 14. Same situation as Figure 13, but
with dual drive motors as the vehicle speed with slip occurring between the road and the
decreases.9 tires.10
The reason for this behavior is that regen torque is a braking operation. And when braking
torque is applied to the rear wheels while slip is occurring as a result of snow, ice, rain, or gravel,
the rear wheels can lock up due to lack of traction. Without rear wheel traction, the vehicle can
become unstable and go into a hazardous spin about its vertical axis that cannot be controlled.
Therefore, government regulations require the vehicle manufacturer to use a certain amount of
front wheel braking to prevent this instability from occurring.

Tesla Regen, Brakes and 10 R. Belt

Sudden Acceleration 1 June 2020
The government regulations are based on a brake force distribution curve as shown in Figure 15.
As any given vehicle is braked, the deceleration applied causes the vehicle to tilt forward,
transferring weight from the rear wheels to the front wheels, changing the horizontal braking
forces on the wheels. As weight is removed from the rear wheels, their traction capability is
reduced, and they tend to lock up sooner at lower deceleration values as the brakes are being
applied. However, transferring too much weight to the front wheels can cause the front wheels
to lock up also. It is possible to calculate the horizontal braking forces on the front and rear
wheels at which the front and rear wheels will lock up at the same time as a function of the
deceleration value, the vehicle’s mass, center of mass location, wheel base, and the friction
coefficient of the road. If we plot these values as a curve in a plane with the front and rear
braking forces as the orthogonal axes, we get the ideal I-curve shown in Figure 15. Above the I-
curve the rear wheels will lock first, which is an unsafe condition expressly forbidden by the
government regulations. Below the I-curve any operating pointa are considered safe except for
points below an M-curve which defines the minimum rear braking force for the vehicle. The
slanted lines show the deceleration values for each point along the I-curve. The -curve shows
the operating points associated with the linear brake proportioning valve used in most ICE
vehicles on the road today.

Figure 15. A typical brake force distribution plot used to determine

allowable front and rear braking forces as a function of vehicle deceleration11,12
Using Figure 15, we can now explain the regen behaviors of the Tesla Model 3 vehicle shown in
Figures 13 and 14. The Figure 13 regen behavior is shown as point A in Figure 15, lying on a line
having a deceleration value of 0.3 g. At point A all braking is provided by the rear wheels only,
which puts it deeply in the unallowable region where the rear wheels will lock up first. This is
tolerable only if the the rear wheels have 100% traction on the road, or no slip. If any slip is
present, then to keep the vehicle stable at the same deceleration value, the rear wheel braking
force must be reduced while increasing the front wheel braking force in order to move the
operating point to point B on the I-curve (or below) in Figure 15. At point B the front and rear
braking forces are roughly equal at 0.2 g and 0.1 g, respectively. This corresponds to the regen
behavior of the Tesla Model 3 with slip present as shown in Figure 14. (Figure 15 is not specific
to the Tesla Model 3, so quantitative agreement with Figure 14 should not be expected).

Tesla Regen, Brakes and 11 R. Belt

Sudden Acceleration 1 June 2020
We will now explain how this regen force distribution function is done in Tesla’s traction control
system. Figure 16 shows the traction controller as part of Tesla’s larger drive motor control
system. The vehicle torque command generation function contains the accelerator pedal map
shown in Figure 8 that converts the accelerator pedal value into total vehicle torque request.
The Optimal Torque Split Function contains two drive motor torque maps that convert the total
vehicle torque request into two pairs of motor torque and flux commands, one pair for each of
the front and rear motors, as shown in Figure 9. The optimum torque split is one that runs the
motors at their most efficient operating points in order to conserve battery power, as shown in
Figure 11. This may involve using only one motor at some operating points, however, which
provides acceptable vehicle stability only cases where wheel slip is not present. In cases where
wheel slip is present, the traction control system redistributes the two torques from the
optimum torque split function into two new front and rear torques in order to operate on or
below the I-curve in Figure 15. This achieves the desired vehicle stability.
Two more things must be noted about Figure 16. First, it provides front-to-rear torque
redistribution for traction control purposes under both acceleration and regen deceleration.
However, it does not provide more general differential (i.e., lateral) wheel slip control for other
vehicle stabilization purposes such as oversteer or understeer. These more general wheel slip
control functions are provided by the vehicle’s braking system. Second, the vehicle torque
command generation function has an input from the vehicle stability control system (VSC). This
is unexpected, but is not an error, as it appears in four of Tesla’s ten patents on this system.
This seems to give the VSC system authority over the vehicle’s motor torque control, in addition
to the accelerator pedal. More will be said about this input in the next section on Tesla’s braking
system.

Figure 16. Tesla’s drive motor control system.13 Computed signal values C_nnnn are 8-bit
sign-magnitude integers in the Model 3 and 16-bit sign-magnitude integers in the Models
S and X that are recomputed every 10 ms (100 Hz). Redundant signals are not used.

Tesla Regen, Brakes and 12 R. Belt

Sudden Acceleration 1 June 2020
Figure 17 shows a block diagram of Tesla’s traction controller function.Note 1 The controller
consists of four sections. The first section contains the first stage of traction control which
independently minimizes wheel slip using a PID-based feedback control system.

Figure 17. Block diagram of Tesla’s traction controller.14 Computed signal values C_nnnn are 8-
bit sign-magnitude integers in the Model 3 and 16-bit sign-magnitude integers in the Models S
and X that are recomputed every 10 ms (100 Hz). Additions in red are explained in footnote 1.
When wheel slip occurs on either the front or rear wheels, the torque on that axle is reduced and
transferred to the other axle that has a lower wheel slip. This is seen from the following
equations, which are obtained from Figure 17:
C_ torque1 = (1- 1) C_torque1e + K1 • (C_ torque − (1- 1) C_torque1e − (1- 2) C_torque2e)
C_ torque2 = (1- 2) C_torque2e + K2 • (C_ torque − (1- 1) C_torque1e − (1- 2) C_torque2e).

Note 1 This block diagram, which appears in ten of Tesla’s patents, shows summing operations in four
places that are clearly incorrect because the two inputs have different units (adding apples and oranges
can’t result in more oranges). Instead, these four summing operations should be understood as
modulation operations (i.e., multiply operations), in which the torque T is pulse width modulated i.e.,
multiplied) by a duty cycle  between 0 and 1 that results in a reduced torque of magnitude (1-  times T.
The duty cycle  is determined by a PID controller that drives the current slip ratio to a target slip ratio
that is either zero for a straight drive path or some normal minimum slip ratio for a curved drive path,
where the minimum slip ratio can change with vehicle speed and steering angle (wheel speeds normally
vary during a turn even with no slip). The slip ratio is given by the difference between the wheel speed
and the vehicle speed divided by the larger of the two. The input to the PID controller is the slip error,
which is the difference between the current slip ratio and the target slip ratio, which is obtained from a
look-up table. The PID controller drives this slip error to zero, which drives the torque from some
reduced value (1-  T when wheel slip is present to the full value T commanded when no wheel slip is
present.
Tesla Regen, Brakes and 13 R. Belt
Sudden Acceleration 1 June 2020
With no slip on either axle, 1 = 0 and 2 = 0, and we get:
C_ torque1 = C_torque1e + K1 • (C_ torque − C_torque1e − C_torque2e)
≈ C_torque1e
C_ torque2 = C_torque2e + K2 • (C_ torque − C_torque1e − C_torque2e)
≈ C_torque2e,
As expected, this shows that the output torques C_ torque1 and C_ torque2 are merely the same
as the input torques C_torque1e and C_torque2e, respectively.
If we have maximum slip on axle 1 and no slip on axle 2, then 1 = 1 and 2 = 0, and we get:
C_ torque1 = K1 • (C_ torque − C_torque2e)
≈ K1 • C_ torque1e
C_ torque2 = C_torque2e + K2 • (C_ torque − C_torque2e)
≈ C_torque2e + K2 • C_ torque1e
In this case, the output C_ torque2 with no slip is increased from C_torque2e to C_torque2e +
K2 • C_ torque1e, which transfers torque from axle 1 to axle 2. Meanwhile, the output C_
torque1 with slip is reduced from C_torque1e to K1 • C_ torque1e, which reduces the slip but
provides a dynamic boost component K1 • C_ torque1e.
In the second stage of traction control (third section) motor speed fast disturbances are
independently minimized using a high pass filter and a second PID-based feedback
controller. Motor speed fast disturbances can be caused, for example, by sudden large
reductions of load torque on the motor shaft during an excessive wheel slip event, or by sudden
large additions of load torque on the motor shaft from one or two stuck wheels.
Between the first and second stages is a transient torque boost feedforward control circuit,
referred to in the figure as dynamic boost, which adds an amount of torque to each axle. The
amount of torque added is proportional to the difference between the driver torque request after
the first stage of traction control and the combined torque command, C_torque. The
proportional constants K1 < 0 and K2 < 0 can be tuned to be different values for the two axles.
The feedforward torques enhance the vehicle performance, vehicle response to driver request
and drivability without compromising traction control and vehicle stability. The feedforward
torques are zero when the torque request is fully met, with zero effective wheel slip ratio errors
and with the maximum torque limits not in effect. During a wheel slip event that causes a
torque reduction on an axle, an effect of the feedforward control is to increase the torque
command to the other axle that has a better tire-to-road grip. The feedforward control also adds
a torque command to the axle experiencing wheel slip, but due to the relatively smaller gain in
the feedforward path, the wheel slip ratio error feedback loop still dominates and will minimize
the wheel slip ratio error.
After the second stage of traction control, in the last section, torque limiters independently limit
the torque commands issuing from the second stage based on C_maxtorque1 and
C_maxtorque2. This stage assures operation in the the regen portion of the motor torque map
of Figure 10 when the torque is negative. The output of the torque limiters are torque
commands C_torque1 and C_torque2.
This controller only provides traction control in response to longitudinal wheel slip that can be
minimized by front-to-rear torque redistribution It does not provide stability control functions
based on lateral wheel slip that are solved by left-to-right torque redistribution, like oversteer
and understeeer. These additional stability control functions must be supplied by the vehicle’s
braking system.
Before discussing Tesla’s braking system, it must be mentioned that some of Tesla’s vehicles,
including the Model 3 described in Section 1, have only a single rear drive motor. In this case,
regen torque can’t be transferred from the rear drive wheels to the front drive wheels because

Tesla Regen, Brakes and 14 R. Belt

Sudden Acceleration 1 June 2020
there is no front drive motor. Yet, the vehicle stability requirements of Figure 15 must still
apply. The only solution in this case is that some of the regen torque is transferred from the rear
drive axle undergoing slip to the front wheel friction brakes. This may sound to some readers
like brake blending, which others have argued that Tesla does not do. More will be said about
this in the next section.
IV. Tesla Braking System
The service braking system used on all Tesla vehicles consists of three major components:
1) a brake booster with its associated electronic control module containing a brake force
distribution function,
2) a brake modulator unit with its associated electronic control module containing slip
control functions for ABS and ESC.
3) four wheel assemblies that contain the disc brake actuators and brake pads along with
wheel speed sensors.
There is also a completely separate parking brake system with its own controller and brake
actuators and pads that will not be discussed here.
The brake booster used in Tesla’s brake system is Bosch’s iBooster as shown in Figures 18 thru
21. It uses an electric motor to provide brake pedal assist or boost instead of a large circular

Figure 18. Bosch’s 1st gen iBooster with Figure 19. Cross section of the 1st gen iBooster
tandem master cylinder.15 showing the electric motor and gear drive. 16

Figure 20. Bosch’s 2nd gen iBooster Figure 21. Cross section of Bosch’s 2nd gen
used in Tesla vehicles.17 iBooster showing the three-stage gear unit.18

Tesla Regen, Brakes and 15 R. Belt

Sudden Acceleration 1 June 2020
 Figure 22. Cross section of a conventional vacuum booster
used on most vehicles on the road today
vacuum chamber as used in the conventional vacuum booster shown in Figure 22. Besides the
electric motor that provides the power boost, the Bosch iBooster contains a boost motor
controller with PID feedback control and high power drive transistors, a tandem master cylinder
that supplies brake pressure to the front and rear brake actuators, two brake pedal travel
sensors that enable dual redundant sensor outputs, and a programmable electronic control unit.
It also uses an external brake lamp/STOP switch to sense brake pedal depression for purposes of
brake light activation and control of related vehicle functions. The control unit translates the
driver’s force on the brake pedal into hydraulic pressure on the brake fluid as well as electronic
commands to the solenoid valves in the brake actuator module, enabling them to control the
transfer of the brake fluid from the master cylinder to the front and rear brake actuators with an
appropriate brake force distribution function. For safety considerations, the iBooster is
designed to allow the driver to apply the brakes manually if boost power is lost for any reason.
Although the iBooster is used in Tesla vehicles for these purposes, it is not essential for them
because the same functions could be provided by a conventional vacuum booster. This shows
that the iBooster is desired for other reasons, as discussed Appendix I.
The brake modulator unit used in Tesla’s brake system is Bosch’s ESP hev II module as shown in
Figure 23. It is a special version of Bosch’s newest ESP 9.0 ABS modulator that is specifically
adapted for use in hybrid electric vehicles. Figure 24 shows that the unit consists of over twelve
electrically operated solenoid valves of various types, two hydraulic pressure pumps, an electric
motor to operate the pressure pumps, a PID controller with feedback control and high power
drive transistors to operate the electric motor, accumulators, pressure sensors, and an overall
electronic control module for issuing commands to the pumps and solenoid valves. The control
module receives commands from the iBooster via a high speed serial CAN bus. Response time
for commands is one millisecond to enable fast braking operations in emergency situations.

Tesla Regen, Brakes and 16 R. Belt

Sudden Acceleration 1 June 2020
 Figure 23. Bosch’s ESP hev II Figure 24. Expanded view of Bosch’s ESP hev II unit
brake modulator unit.19 showing its internal components.20
Figure 25 shows how the iBooster and the ESP hev II brake modulator work together to achieve
brake operation. The valves are shown in their normal operating positions when power is off,
which allows the transfer of pressurized brake fluid from the master cylinder to the wheels to
allow normal manual operation of the brakes. Besides mechanical activation of the brakes by
the brake pedal, the system also allows electrical activation independently of the brake pedal by
closing the isolation valves from the master cylinder and activating the pressure pumps to
provide the desired brake pressure to the wheels. This electrical activation can be done on either
channel separately while the other channel is operated manually or on both channels at the
same time. When electrical activation is used on either channel, the driver can feel the brake
pressure provided by the electrically activated pressure pump as it is fed back into the master
cylinder via a pressure relief valve, as shown by the right hand channel in Figure 25. In hybrid
electric vehicles, it is often desireable to prevent the driver from feeling the electrically activated
brake pressure in order to hide changes in the manually operated brake pressure caused by
brake blending operations associated with regen. Therefore, Bosch has modified its existing
ESP 9.0 modulator for this purpose by returning the brake fluid from one channel back to the
fluid reservoir instead of the master cylinder as shown on the left channel in Figure 25. This
prevents the driver from feeling the electrically activated brake pressure in this channel. This
also requires a special kind of pressure relief valve as shown in yellow in Figure 25. Together,
these ESP hev II modifications support brake blending operations in a hybrid electric vehicle.
Tesla vehicles use the ESP hev II unit with the left-hand channel on the rear wheels and the
right-hand channel on the front wheels Note 2. This allows the left-hand channel to electronically
blend (i.e., substitute) friction brake pressure with regen brake pressure in any amount up to 0.3
g, as limited by the maximum pressure provided by the electric pump in the modulator unit.

Note 2. Most non-electric vehicles use a so-called diagonal or X-split, configuration in which one channel
activates the left front brake and the right rear brake while the other channel activates the right front
brake and the left rear brake. This is done to maintain at least one front brake for additional safety in case
one channel is lost. Hybrid electric vehicles, on the other hand, prefer the so-called parallel or front-rear
configuration for ease of implementing blending operations between friction brakes and regen braking.
Tesla Regen, Brakes and 17 R. Belt
Sudden Acceleration 1 June 2020
Figure 25. Hydraulic valve system diagram showing how the valves in the ESP hev II module
work together with the iBooster module. Valves are shown in their normal operating positions
when power is off, which allows the transfer of pressurized brake fluid from the master cylinder
to the wheels.21
One operation the modulator in Figure 25 cannot support is the transfer of regen braking force
from the rear motor into friction braking force on the front wheels while the driver is applying
friction braking force to the front wheels. This is essential in a vehicle having only one rear drive
motor such as the Tesla Model 3. Therefore, Bosch has added another modification to its ESP
9.0 modulator in the form of an electronically operated plunger as shown in Figure 26. The
plunger uses an electric motor to push a piston that forces brake fluid out of a storage chamber,
thereby increasing the brake pressure just like the master cylinder increases brake pressure in
response to a piston pushed by the driver. This modification is not needed on dual drive motor
vehicles, so it is unknown whether all Tesla vehicles use this modification or whether vehicles
with only rear drive motors do, because it does increase the cost of the modulator.

Tesla Regen, Brakes and 18 R. Belt

Sudden Acceleration 1 June 2020
Figure 26. Variation of Bosch’s ESP 9.0 modulator shown in Figure 25 that adds a plunger for
electrical activation of the front brakes while the front brakes are being activated manually.22
By using one of the ESP modulator units shown in Figures 25 or 26 it is possible to do brake
force distribution of the friction brakes when the brake pedal is applied that meets the
government requirements shown in Figure 27. In this case, the iBooster provides the
instructions to the ESP modulator on how to distribute the total vehicle braking force to the
front and rear wheels, and the ESP modulator carries out these instructions by suitable control
of the modulator solenoid valves and modulator pressure pumps. Instructions are provided via
a high speed CAN bus. Instructions can be devised to provide any given deceleration path in the
unshaded (white) portion of Figure 27. The dashed blue line is one example of such a path. In a
vehicle with rear wheel drive, this path is likely to follow the ideal I-curve shown in Figure 27.
The distance from the origin to any point along the path gives the deceleration value z (also
known as the braking strength or braking severity), which is proportional to the TMC pressure
or the total braking force. This brake force distribution operation is distinguished from the
brake force distribution of the regen braking torque in Figure 15 by applying to the total braking
force when the brake pedal is pressed (i.e., friction brakes as well as regen braking), instead of
applying to regen braking only when no brake pedal is pressed as shown in Figure 15.

Tesla Regen, Brakes and 19 R. Belt

Sudden Acceleration 1 June 2020
 Figure 27. Brake distribution function for the total brake force (friction plus regen).23
The dashed blue line shows an arbitrary distribution function possible using the iBooster.
Friction brake application, either manual or electrical, with the proper brake force distribution
between the front and rear wheels, is only half of the functions performed by the ESP hev
modulator unit. The other half of the functions are slip control functions that provide electronic
vehicle stabilization. These slip control functions include the following:
1. Anti-Lock Braking System (ABS). The ABS controller constantly monitors and
compares signals received from four speed sensors located at each wheel. When the brake is
applied and a wheel is about to lock, the ABS modulator reduces the brake pressure at the
unstable wheel enough to prevent wheel locking. As soon as the wheel becomes stable, the
ABS controller increases brake pressure, thereby maintaining optimal braking force at all
times. The ABS control system performs these adjustments several hundred times per
second on each wheel independently. This enables the driver to achieve the highest brake
force physically possible, while maintaining vehicle stability and steering over any road
surface.
2. Dynamic Traction Control (DTC). DTC prevents excessive motor torque from reaching
the driven wheels. By monitoring wheel speeds, the TCS controller in the brake system can
modulate and reduce motor torque on only one driven wheel at a time by braking the
slipping wheel to keep it from spinning. The operation is similar to ABS, but takes place
during acceleration instead of deceleration as with ABS. This function provides improved
traction control when only one side of the car is driving over a low traction surface, such as
patchy ice, where only one driven wheel has low grip. Therefore, DTC augments the
operation of the traction controller in Tesla’s drive motor control system, which removes
torque from both wheels on one axle at the same time, by imitating the function of a
conventional differential lock. This function is especially useful when accelerating or driving
up a hill on a loose surface or in deep snow.
3. Dynamic Brake Control (DBC) DBC assists the driver in panic or emergency braking
situations by automatically boosting the brake pressure. In emergency braking situations,
drivers often fail to apply sufficient pedal force to achieve the maximum vehicle deceleration.
ABS regulation is then not activated. The DBC control unit monitors inputs from the brake
light switch and the brake pressure sensor. The trigger for activation of DBC is how rapidly
the brake pressure is increasing with an application of the brake pedal. If the trigger
condition indicating an emergency braking event is satisfied, DBC uses the ABS pressure

Tesla Regen, Brakes and 20 R. Belt

Sudden Acceleration 1 June 2020
 pump to increase brake pressure when the driver’s pedal force is insufficient. Hydraulic
brake pressure is artificially increased much faster than pressure from the driver, and is
maintained at the optimal ABS operating pressure. This pressure increase is done in the
following situations: a) when the brake pedal is rapidly depressed with insufficient pedal
force, or b) when the brake pedal is depressed slowly and the demand for deceleration is
subsequently high, after one wheel reaches the ABS control threshold. This function is
useful when the traffic slows, making light braking necessary at first, but then something
happens that demands as short a stopping distance as possible. It is active at any speed
above 3 mph. This function is sometimes called Dynamic Brake System (DBS) or Brake
Assist System (BAS).
4. Automatic Emergency Brake (AEB). AEB decelerates the vehicle without the driver
having to use the brake pedal when the probability for forward collision is predicted as high.
The purpose of AEB is to eliminate or mitigate collisions where subject vehicle collides with
a lead vehicle. AEB uses the friction brake system as actuator, up to full brake which would
be typically 10 m/s2. An AEB system must be designed to not trigger too early, because the
driver could be disturbed by it, thereby actually causing accidents. Therefore, in many
situations, AEB will mitigate collisions rather than avoid them
5. Corner Brake Control (CBC). CBC can activate if the vehicle is cornering while ABS
regulation is not taking place. When a transverse acceleration in excess of 0.6 g is detected
while the brakes are being applied, the CBC prevents a buildup in brake pressure to the
inside rear wheel or the outside front wheel. The difference in braking force between the two
wheels on the same axle creates a yaw force that opposes the transverse acceleration and
allows the vehicle to handle neutrally. The CBC controller accomplishes this by closing the
inlet valve, thus not allowing brake pressure to increase at the brake caliper of the desired
wheel. This attempts to prevent the vehicle from entering into an unstable situation that can
lead to understeer or oversteer. This function only works while the brakes are being applied.
6. Electronic Stability Control (ESC). ESC maintains control of the vehicle during
cornering. It uses yaw rate and acceleration sensors in addition to wheel speed sensors. ESC
calculates the intended path of the vehicle using inputs from a steering angle sensor and
compares it to the measured rate of turn of the vehicle from the yaw rate sensor. This allows
it to monitor for understeer or oversteer events during cornering. In the case of under-
steering, the rear inside wheel is braked, resulting in a positive yaw torque that helps the
vehicle turn into the corner. During oversteering, the front outside wheel is braked resulting
in a negative yaw torque that steers the vehicle out of the corner and helps the rear axle
regain traction. ESC also controls and limits engine power to the extent necessary to support
lateral tire grip during cornering. This function works when the brakes are not being applied
or when the lateral acceleration is less than 0.6 g.
7. Engine Drag Torque Control (EDC). EDC prevents deceleration torque (i.e., drag
torque) from the engine from locking the rear drive wheels, which can generate wheel slip
that can cause the vehicle to spin about its vertical axis when decelerating on roads having
low friction. The situation is similar to deceleration caused by excessive brake application
on a low friction surface, which would normally cause activation of the ABS system that
shifts the operating point on the brake distribution curve to one closer to the ideal curve.
However, with engine drag causing the deceleration, reducing the deceleration by lowering
the braking force using ABS is ineffective because the braking torque is not caused by the
friction brakes. Instead, it is caused by the negative drag torque of the internal combustion
(ICE) engine, which may be further increased by shifting into a lower gear. The only way of
reducing this negative engine drag torque is to increase the engine torque by sending a
throttle request to the engine for a higher positive torque when slip is detected that can
cause the rear wheels to rotate slower than the front wheels with no brakes being applied. In
electric vehicles without an ICE engine, this negative drag torque is produced by the regen

Tesla Regen, Brakes and 21 R. Belt

Sudden Acceleration 1 June 2020
 torque of the electric drive motor, which causes a deceleration of about 0.2 to 0.3 g’s. If this
regen deceleration is applied suddenly on a slippery road, it can cause an unstable situation
that requires this function. In this case, the EDC controller sends a torque request to the
drive motor to provide a positive torque that will offset, or reduce, this negative regen torque
for a short period of time. Since the EDC controller does not know the magnitude of this
torque, it can only keep incrementing the torque request in the positive direction until it
senses that the slip on the drive axle is zero. This function is provided by the ABS unit in all
ICE vehicles and electric vehicles. It is often referred to as MSR, the German acronym for
Motor Schleppmoment Regelung (MSR), which is the German translation of Engine Drag
Torque Control.
The last drag control function may strike the reader as particularly interesting because it gives
the ESP hev II module authority over the drive motor torque. This explains the input to the
drive motor torque controller from the Vehicle Stability Control system shown in Figure 16. But
this authority should not be enough to cause sudden acceleration because it is limited to
increasing the motor torque by only 0.2 to 0.3 g’s to offset a negative regen torque to make it
equal to zero. Nevertheless, this authority over drive motor torque necessitated a further look at
patents to determine how this function operates. This study found several patents that added
important details to the explanation of this function. US patent 6535809 by Delphi explained
that the brake switch is checked to determine whether the cause of the negative torque is the
engine regen torque and not the friction brakes.24 German patent DE10238224B4 by Bosch
revealed that after the wheel slip has been neutralized, the MSR control is continued for a
predetermined overrun time, and this overrun time is longer when negotiating a curve than
when driving straight ahead, being preferably between one and three seconds.25 Finally, many
patents were found that explained that MSR is more sensitive while cornering than while driving
straight ahead.26,27,28 We will have more so say about this function in Section V.
All of these slip control functions are included in the ESP hev II modulator as software built into
the modulator control unit by the brake system subcontractor, who in this case is Bosch. This
software is proprietary to the modulator subcontractor and not accessible for modification by
the vehicle manufacturer. The software executes under the local control of the ESP hev II
module control unit acting as a master on the CAN bus, and the functions are always on. This is
contrary to the brake force distribution function, which operates as a slave to the iBooster
controller and is active only during manual braking. Although some vehicle manufacturers give
the driver the ability to turn off some of these stability control functions, Tesla does not. This
makes it impossible for most Tesla drivers to “do dougnuts” or to do tighter turns with induced
slipping.Note 3
We will now look further into the operation of Tesla’s braking system and how its ESP hev II
module interacts with the regen system by examining the requirements that government
regulations impose on them. Further information on the stability control functions is obtained
from patents and technical papers.
V. Further Information on Tesla’s Braking System and Its Interaction with Regen
As Tesla vehicles are now being sold internationally, they are subject to government
regulations.Note4 These regulations specify how the braking system must operate and how its

Note 3. Some Tesla drivers have found a way to turn off these functions by activating a secret
dyno mode, and Tesla has provided a new Track Mode on Model 3’s to do this. See
https://www.teslaincanada.ca/model-3/tesla-has-a-secret-dyno-mode-that-disables-traction-and-stability-control-
and-more/ and https://www.tesla.com/es_ES/blog-/how-track-mode-works?redirect=no.
Note 4. Tesla’s original Roadster did not follow these regulations because it was granted an exception as a
result of being an experimental vehicle. But all subsequent Tesla models must follow these regulations.

Tesla Regen, Brakes and 22 R. Belt

Sudden Acceleration 1 June 2020
ABS modulator functions must interact with the regen system. These regulations are provided
in the following government documents:
1. United Nations Economic Commission for Europe (UN/ECE) Regulation No. 13-H
(2017) with Annexes 1 to 9, “Technical Requirements, Test Methods, and Limit Values
for the Braking Systems of Passenger Cars and Light Commercial Vehicles”.
2. United Nations Economic Commission for Europe of the (UN/ECE) Regulation No 140
“Uniform provisions concerning the approval of passenger cars with regard to Electronic
Stability Control (ESC) Systems”, 2018/1592
3. USA regulation 49 CFR § 571.126 Federal Standard No. 126; “Electronic stability control
systems for light vehicles”.
These regulations are discussed in greater detail in the following government-sponsored reports,
which also serve as secondary source material for the requirements:
4. “ID4EV -- Intelligent Dynamics for Electric Vehicles”, a study on the requirements and
specifications for electric vehicles derived from user needs and SOTA analysis performed
under a European research project co-funded by the Economic Commission for Europe
(ECE) under the 7th Framework programme.
5. “Investigation Into Regenerative Braking Systems” by B. J. Robinson, C. Visvikis, T.
Gibson, and I. Knight of the Transport Research Laboratory, published project report
No. PPR582 for the Department for Transport, Transport Technology and Standards,
United Kingdom, 2011.
6. “Functional Safety Assessment of a Generic, Conventional, Hydraulic Braking System
with Antilock Brakes, Traction Control, and Electronic Stability Control”, by C. Becker,
D. Arthur, and J. Brewer of the Volpe National Transportation Center, DOT report No.
DOT HS 812 574, Volpe report No. DOT-VNTSC-NHTSA-16-08, 2018.
These regulations and their associated documents provide the following requirements:
1. "Electric regenerative braking system of category A" means an electric regenerative braking
system which is not part of the service braking system; (2.17.2).29 (Tesla’s regen braking
system is Category A).
2. The ESC need not operate below 20 km/h, when in reverse, or when the driver has disabled
it.30
3. The ESC shall do a start-up self-test within 2 minutes after vehicle starting.31 (6.2.3)
4. A defective brake light switch (aka STOP lamp switch) can trigger the ESC OFF indicator. A
blown brake light bulb can also trigger the ESC OFF indicator.32
5. Electric regenerative braking systems as defined in paragraph 2.1.7 of this Regulation, which
produce a retarding force upon release of the accelerator control, shall generate the signal
mentioned above (ie, brake light signal or STOP light signal) according to the following
provisions:
Vehicle decelerations Signal generation
≤ 0.7 m/s2
(≤ 0.07 g’s) The signal shall not be generated
> 0.7 m/s2 and ≤ 1.3 m/s2 (0.07 ≤ x ≤ 0.13 g’s) The signal may be generated
> 1.3 m/s2 ( > 0.13 g’s) The signal shall be generated

In all cases, the signal shal be deactivated at the latest when the deceleration has fallen
below 0.7 m/s2.33 (g-equivalents in parentheses added by author). [The maximum regen of

Tesla Regen, Brakes and 23 R. Belt

Sudden Acceleration 1 June 2020
 most Tesla models is 0.2 g’s, which was increased to 0.3 g’s with software update 2018.42
v9].
6. ESC must remain capable of activation even if the antilock brake system or
traction control system is also activated.34
7. The integration of the regenerative brake system controller into the anti-lock braking system
is regulated in paragraph 2.5.18.4. In case an anti-lock braking system is installed, its ECU
must incorporate the control of the regenerative brake system. Deactivation of the
recuperation by an automated system is explicitly permitted. Therefore an electronic
stability control system can deactivate the regenerative braking if required. In the case of the
regenerative braking system being linked to the anti-lock brake system, it is subject to the
test procedures defined in annex 6 – test requirements for vehicles fitted with anti-lock
systems.35
8. 12 V DC power supply network. In case ABS, ESC or such like systems apply corrective
actions (e.g. in cases of excessive wheel slip, under steer or over steer), the recuperative
brake system is deactivated and only friction brakes are used. State of the art brake system
controllers do not adapt recuperative brake torque to the driving conditions. 36
9. If ABS becomes active the regenerative torque shall be controlled or switched off until the
end of the brake (BLS = off).37
10. ECE R 13-H (additional sections for regenerative braking system to be fulfilled):
The regenerative brake shall consider the load of the vehicle (chassis level) and the level of
adherence. Furthermore it has to be controlled by the ABS controller to avoid excessive
wheel slip and to guarantee the stability and the steer ability of the vehicle.38
11. ABS activates when the wheels begin to lock. This occurs most easily on surfaces with a low
coefficient of friction such as ice. However, when ABS activates, regenerative braking is
usually switched off to protect the normal ABS function (Zhang et al., 2010). In these
circumstances, the friction brakes would need to compensate for the loss of regenerative
braking in order to maintain the same level of deceleration. The strategy for integrating
regenerative braking with ABS, and in particular, the timing of the transfer to friction
braking, might affect the braking performance of the vehicle and the pedal feel.
Where vehicles with regenerative braking are fitted with ABS, it is a requirement that the
ABS controls the regenerative braking as well as the friction braking in order to enable the
system to reduce the wheel torque in the event of lock. However, there is no requirement as
to how this should be controlled, so in theory it could be modulated in the same way as the
service brake system or simply disconnected. If suddenly disconnected, there is again a
general requirement that the friction brake should compensate for the change in
deceleration that would otherwise resjult from disconnection, which may affect pedal feel.
The Vehicle Certification Agency (VCA), the UK Government agency responsible for granting
type approval to vehicles, interpret this requirement to mean that compensation must be
provided whenever the regen component is disconnected. But an alternative interpretation
is conceivable; namely, that this compensation (friction brake achieving 75% of its final
value within 1 second) is only required where the regenerative braking syhstem alone is
providing braking torque prior to the disconnection. The wording of the Regulation is
considered a little ambiguous in this regard.39
12. ABS. An OEM stakeholder (i.e., manufcturer) suggested that there was unlikely to be any
problems with a regenerative system’s interaction with the ABS system, because the contgrol
algorithms simply switch the regen off as imminent wheel lock is detected by the wheel
speed/slip sensors, and the conventional \braking system, controlled by the ABS, takes
over.40
13. ESC. Few studies have been reported on the interaction between regenerative braking and
ESC and the effects on vehicle stability. Hancock and Assidian (2005) investigated the
impact of regenerative braking on vehicle stability during cornering. A full vehicle model of

Tesla Regen, Brakes and 24 R. Belt

Sudden Acceleration 1 June 2020
 a hybrid sports utilioty vehicle was used for the study (in a computer simulation). They
found that applying regenerative braking to the rear axle can reduce stabilty, depending on
(drive) motor size and f\oad surface friction coefficient. With a moderately sized motor on a
high  surface, the reduction in stability was controlled by the ESCsystem without a large
increase in ESC brake pressures. However, on a low  surface the reductiuon in stability was
much more severe and could not be compensated for by the ESC. Two solutions were
proposed to prevent regenerative braking from causing the wheels to slip; firstly, switching
to friction braking once the longitudinal slip of either rear wheel exceeds a specified
threshold and, secondly, locking the centre coupling. Both solutions were effective from a
stability perspective, but locking the centre coupling had the added advantage of maximizing
energy recovery. Howeve, the authors recommended further work to assess the potential
effects on ABS/ESC performance.
14. ESC with ABS. Hard braking on a low friction surface will activate the ABS. In the critical
test of (hard) braking while cornering, the ESC becomes active. Both occasions will swtich
off the regenerative brake system (RBS) and noe of the brake energy is recuperated. During
these events it becomes more important to retain steerability and vahicle stability than
redover energy.
15. ESC. Braking-in-a-turn tests would be suitable for investigating the stability effects of low
levels of longitudinal deceleration in combination with high lateral acceleration when the
brake ratio is strongly biased to one axle as a result of regenerative braking.
16. ABS must control regeneratove braking. If the ABS wheel speed signal sees excessive slip, it
will trigger suspension of the regenerative braking until the next brake event or lift-off event.
For vehicle systems that use non-zero regenerative braking torque when an ABS evbent
occurs, vehicle instability is observed.
17. It is unclear from the existing literature whether current regenerative braking systems, in
general, remain switched off when ABS has been activated until the next brake application
even if the ABS deactivates during the stop, or if they are simply switched off only when the
ABS is activated, and back on again whenevern it isn’t. The tests conducted for this study
(the low-high transition tests) suggest that the test vehicle’s regenerative braking system is
reactivated once the ABS is deactivated. Some designers specify that the regenerative
braking cannot be used again until the next brake application (e.g., Zhang et al., 2008).
However, Zhang et al., (2010) proposed a system in which regenerative braking is
reactivated when ABS control is ended.
18. Continental markets their regenerative brake system with the statement “The use of a
conventional ESC unit enables the regenerative brake system to perform all known braking
interventions and stability functions.” This essentially brake-by-wire system also has a
direct hydraulic push-through in case of brake-by-wire failure. The pressure applied by the
driver to the brake pedal is measured and converted into a braking torque demand. The ESC
unit is enhanced to communicate with the vehicle’s electric drive motor system (motor +
battery + battery management). The electric motor system reports to the ESC unit how
much regenerative braking torque is available (depending on the state of the charge and
temperature). The ESC unit sends back a braking torque demand to the electric drive motor
system that does not exceed the available regenerative braking torue and then ensures that
the rest of the demand is met by the hydraulic system. There is, therefore, complete
flexibility in control to mix braking from the electrical and friction systems, so it is possible
to limit the braking to the capabilities of the battery but, at the same time, to maximize
energy recapture.
19. According to Economic Commission for Europe (ECE) regulation 13-H, installation of an
anti-lock braking system (ABS) in a vehicle is required and must have the supreme priority
to take control of vehicle braking in any situation. Similar rules have been promulgated by
the United States National Highway Traffic Safety Administration.

Tesla Regen, Brakes and 25 R. Belt

Sudden Acceleration 1 June 2020
20. ABS. The National Highway Traffic Safety Administration (NHTSA) received over 1200
complaints from Toyota Prius owners alleging a momentary reduction in braking
performance on uneven road surfaces (NHTSA, 2010). Toyota received nearly 200 further
complaints directly. The company performed brake tests in an effort to reproduce this
phenomena and found that the braking force reduced after ABS activated. (Toyota, 2010).
Further investigations revealed that many drivers had experienced the phenomenon,
particularly in winter, where drivers may maintain a fixed pedal stroke. Under these
conditions, Toyota found that “vehicle stopping distance may increase, relative to the
driver’s expectations for a given pedal force.” The condition was a result of the ABS
software, which was permitting a change in braking force. Although the ABS was operating
as it was designed to, Toyota conducted a voluntary safety recall to re-program the ABS
control unit in the Prius (and the Lexus HS250H). It appears that, in this particular system,
the transition from regenerative braking to service braking (i.e., hydraulic braking) was
being perceived by the drivers as a change in braking force. Under certain circumstances,
the driver would need to press harder on the brake pedal to maintain the same deceleration.
21. Electronic Braking Function requirements:
1. Proportion brake force between front and rear wheels to maximize braking effectiveness.
2. Proportion brake force between left and right wheels to maximize braking effectiveness.
3. Control brake fluid pressure to prevent vehicle wheels from locking-up under braking
during ABS events.
4. Provide selective wheel braking during Traction Control System (TCS) events.
5. Control brake fluid pressure to each wheel to provide vehicle control during Electronic
Stability Control (ESC) events, including during extreme dynamic maneuvers and in
adverse roadway conditions.
6. Provide brake force to support other advanced braking features (e.g., hill holder).
7. Implement braking requests to support other vehicle systems.
8. Measure and provide the vehicle speed using available sensors and models.27
9. Coordinate yaw rate stabilization with the steering system and other vehicle systems.
10. Communicate with internal subsystems and external vehicle systems.
11. Request an increase in torque from the Drive Motor Controller/Electronic Throttle
Controller to prevent wheel lock during sudden deceleration.
12. Request a reduction in propulsion/throttle from the ACS/ETC when needed to support a
Traction Control System (TCS) or Electronic Stability Control (ESC) event.
Note that requirements 21.11 and 21.12 specify that the braking system can request either an
increase or decrease in torque from the drive motor in response to either a sudden deceleration
or a traction control/electronic stability control event, respectively
In summary, these requirements show that regen not under the control of the friction braking
system is incompatible with the slip control functions of the brake modulator module. This
incompatibilty is true not only for Class B vehicles that use the brake pedal for activating regen,
and must therefore blend the regen torque with the friction braking torque, but also for Class A
vehicles like Tesla vehicles that use the accelerator pedal for one pedal driving. This
incompatibiility arises because the slip control algorithms in the braking modulator need to
know all the braking forces acting on the vehicle at the time the algorithms are active, and lose
their effectiveness when the braking forces caused by regen are not under their control. This is
why the regulations require that the braking system must be in control of all braking forces,
including the braking forces caused by regen. And this is why regen must be turned off when
these algorithms are activated, because turning regen off reduces the forces on the vehicle to the
normal case for braking system forces only. This is the case for all vehicles having internal
combustion engines, which is the case assumed when the algorithms were originally developed.

Tesla Regen, Brakes and 26 R. Belt

Sudden Acceleration 1 June 2020
 VI. Loss of Regen While Going Over Bumps
Loss of regen while going over bumps happens in all electric vehicles, whether they are hybrid
vehicles or battery-powered vehicles only. This loss of regen happens because regen is turned
off by the ABS modulator to prevent forces on the vehicle not controlled by the braking system
from interfering with the operation of the braking system’s algorithms. A good example of this
is found in several discussions of regen found on Tesla forums.
On one forum, the following discussion took place:41
Crebelo
I've had my Model S 75 for 3 months now. Yesterday, I was approaching a red light with no
one in front of me. I was far enough away that I let off the accelerator and let the regen kick
in and slowed down. Then, when getting closer to the stop light, I started hitting the brakes
(I drive like this every day with usually no issues). I hit a large bump in the road and for a
second it felt like the car lost all regen and went into neutral. I was still on the brake pedal
at this time but it definitely felt like the car went from slowing down to speeding up. … This
was the third time it's happened to me, all when coming to a stop and hitting a bump on the
road. … It scared the crap out of me and I know for a fact I was hitting the brake.
jordanrichard
I have had it where I hit a large bump and for a split seconds the traction control light came
on. Any disruption to traction, 0r to wheels touching the ground or wheels stopping from
moving, will cause weird things to happen. You know you hit a bump, but the car doesn't. It
just knows that the wheels stopped moving & lost traction.
murphyS90D
Happens in my 2013 Ford Fusion Energi FWD when one front wheel goes over a sunken
manhole cover. … If either wheel is off of the ground, forward drive or regen braking is not
possible because of the way a differential works. If your foot is not on the brake ABS is not
going to intervene and stop the wheel that is off of the ground. It does feel like acceleration,
but it is not. It is instantaneous removal of regen braking that makes it feel like
acceleration.
Stiction
I used to go over a small speed hump at work under heavy regen and the rear wheels would
chatter.
On another Tesla forum, the following discussion took place:42
gabeincal
I was in San Jose, getting off the freeway, [in preparation for] taking a left turn from the off
ramp. I was doing about 25-30 mph when I got to the intersection. Cruise and autopilot
were off. I had a green light, so crossed the intersection, and while crossing, started to
brake. Hit a couple of uneven potholes in the intersection (it was a little rough to cross).
While braking I felt a short bit of acceleration, like I left the cruise on or something similar,
except my right foot was firmly (and only) on the brake pedal. … I would swear that the car
accelerated by about 5 mph in a very short period of time (probably a few tenths of a
second). It felt like somebody punched the accelerator for me.
McRat
On many cars with regen and ABS, the ABS takes precedent. So when a wheel RPM drops
rapidly without the others dropping, it shuts off the regen and kicks the wheel into ABS
mode. It feels like the car is jumping forward, when all it really is, is the regen shutting off
to allow the ABS to have full control of traction.
gabeincal
Tesla Regen, Brakes and 27 R. Belt
Sudden Acceleration 1 June 2020
 And that is what I was thinking. The ABS stopped the regen. This is possible, although I felt
acceleration, rather than the lack of deceleration. And also, I felt no ABS 'play' on the brake
pedal and no ABS or stability light came on ... But the potholes really might have something
to do with what I experienced.
Logan5
I had a similar experience a few days ago. I was not coming off a highway, just a on a local
road with autopilot off. I noticed I went over some uneven road when this incident
occurred. .. Speed was about 20 MPH at the time.
McRat
You will often feel that 'kick' when you go over a wet manhole cover while decelerating, or
over train tracks at an angle, or when two of your tires are in the middle of the lane where
the oil/gravel normally exists. Wet leaves can do it, too.
Skipdd
I had the same thing happen and I posted about it. … Like you, I called it into Tesla and
gave them the time of occurrence. They confirmed it was ABS. Definitely freaked me out.
Finally, on a third Tesla forum the following discussion was found:43
EVSteve
For perspective, in a Volt an even more serious issue exists … Applying the brakes first
engages the regen braking, then progresses towards friction braking. … The Volt disengages
all regen braking if the system detects a skid or low traction. The gap between the loss of
regen and application of the friction brakes is MASSIVE. The stopping distance in snow and
ice is severely compromised and unpredictable. The only way to avoid this is to disable the
regen system entirely by shifting to neutral and relying on the friction brakes.
Skotty
I think the correct solution is to have the anti-lock braking system also have control over
regen. I would assume they already do in most cases. Hybrids have been using regenerative
braking for many years, and should be subject to the same theoretical problem.
jerry33
Hybrids turn regen off whenever any of the braking safety systems engage (TC, VSC, ABS,
etc.). I believe the Leaf does this, too.
These three discussions show that all Tesla vehicles can experience a loss of regen while going
over bumps in the road, causing a feeling of momentary acceleration or lurching. This loss of
regen is not a problem of poor blending of the brakes with regen. It is a problem because the
vehicle and the user feel the total force on the vehicle at any given time. And when the total
force changes because regen is lost, the vehicle and driver feel will feel this loss of regen as an
acceleration or lurch. This change in deceleration is caused by the ABS system turning off regen
when one wheel speed drops rapidly without the others dropping. It is not the result of a design
decision made by Tesla, but of the algorithms in the ABS modulator that were designed by
Tesla’s brake supplier, Bosch. And Bosch designers had no choice but to turn off regen when
their algorithms activate, because in order for their algorithms to operate properly, they must
have control over all the braking forces on the vehicle while the algorithms are active. And,
having no control over the regen magnitude, they must shut off regen while they are active. This
is true for all other electric vehicles with regen also.
VII. Loss of Regen While Cornering
Loss of regen while cornering also happens in all electric vehicles, whether they are hybrid
vehicles or battery-powered vehicles only. The reason is the same as with loss of regen when
going over bumps. It happens because regen is turned off by the ABS modulator to prevent

Tesla Regen, Brakes and 28 R. Belt

Sudden Acceleration 1 June 2020
forces on the vehicle not controlled by the braking system from interfering with the operation of
the braking system’s algorithms. A good example of this is found in a discussion of the regen
operation of the Ford Fusion hybrid found on the Ford Fusion Hybrid Forum. There, we find
the following discussion Note5: 44
ACDii
“Slowing down from 35 or so and making a right hand turn, when on regen with the regen
indicator on, part way through the turn the car feels as though it is slipping on ice. The
regen stops for a brief moment … Pretty sure it is tied into the steering position sensor …
When the regen braking acts normal the car slows as you step on the brake, with it slowing
more the more you push the brake pedal. The amps are around -45 and dropping down as
the car slows until I come to a stop at 100% … When it is not acting correctly, and I press on
the brake, the car slows a little, the regen indicator on, and the amps to the HVB are in the
mid -50's, near -60. Yet, the car is not slowing down as it should, and I have to press harder
on the brakes to slow it, engaging the pads instead before the car slows. There are times
where I nearly miss my turn since the car is not slowing as expected … Ford says the brakes
[regen?] going away during a turn is normal … I explained in detail that when slowing
through a turn in regen only, when the steering wheel reaches the 90° mark past horizontal
(1/4 turn), the regen stops braking for a brief instant making it feel like the car surges
forward … The problem can occur at any speed. Pulling into my driveway at under 20 MPH
the car will skip when it happens. My driveway is gravel, so the sudden boost of the brakes
going away is quite noticeable ... I mostly notice it when turning off a highway onto another
highway, and all on right angle turns. There is one road that goes from 55 down to 40 and
turns 90° left, then about a mile or so at 40 with a right 90° turn, and it happens at both
turns regardless of speed. Of course, the faster the turn the more I notice it … I did a reboot
of the car, and while the brakes are now working like they should in a straight line braking,
there is still the issue while turning.”
In response, Sleddog writes:
I can confirm there is definitely an issue with the regen brakes ... Pretty sure it is tied into
the steering position sensor ... braking while turning. I can reproduce it, but only at speeds
of less than 10 mph in a turn with light pressure on the brake pedal in regen mode when
mechanical brakes are not active … The drive system seems to be applying more torque to
the inside wheel in the turn, over-powering the regen braking.
Then hybridbear adds some great observations44:
I measured this on my car [a Ford Fusion hybrid] … [Figure 28] shows a chart of the vehicle
speed and traction motor torque when this loss of regen happens while turning left. This
happens on smooth road surfaces, so the loss is not due to bumps causing the loss of regen.
I maintain steady [brake] pedal pressure throughout this whole braking event. You can see
the motor torque [regen torque] drop off. This is when I feel a surge … The regenerative
braking will cut out under specific circumstances and the car will feel like it’s lurching
forward instead of slowing down. The situation is: braking using only regen brakes; no
friction brakes; steady pedal pressure; going 15-20 MPH; turning the steering wheel at least
90 degrees left or right; driving on smooth pavement. Data I have logged from the
Secondary On-Board Diagnostics Module C (SOBDMC) indicates that the traction motor
briefly reduces its negative torque, which causes the car to surge forward. In one instance
that I logged, the negative torque shows around 48 ft-lbs and then suddenly drops to about
8 ft-lbs. The traction motor then gradually increases the torque back to the appropriate
level. In another instance, the negative torque suddenly went from about 28 ft-lbs to 8 ft-

Note 5. Extracts from several messages by the same contributor have been joined by ellipses in order to
save space and to keep the discussion on topic. This has also been done for other contributors as well.
Tesla Regen, Brakes and 29 R. Belt
Sudden Acceleration 1 June 2020
 lbs. The drop in torque occurs in 1/10 of one second per my logged data. It takes about 2.5
to 3.0 seconds for the traction motor [regen torque] to return to the appropriate level of
negative torque to stop the vehicle.
How sharp does the turn have to be and how fast to notice the problem? … I'm not so sure
that the steering wheel angle is all that important. I think it's more about the difference in
wheel speed [Figure 29] I notice it most often when making left turns & slowing down,
such as when turning into a parking lot or driveway. The lurch seems to happen between 15-
20 MPH.

Figure 28. When the steering wheel is Figure 29. When the steering wheel is
turned 90°, regen is lost and a forward turned 90°, the wheel speeds change and
surge is felt.44 regen is lost, causing a forward surge.44
When braking and turning, the "operating as designed" behavior appears to be when the
difference in wheel speed reaches a certain threshold. The car then reduces regen braking
& blends in the friction brakes. This transition should be seamless & imperceptible. Since
our car had the 14S21 & 15E03 recalls completed, this transition has become perceptible.
This is the issue. The car is "operating as designed" to reduce regen while turning. The
problem is that there is now a momentary delay between when regen is reduced and when
the friction brakes are engaged. There shouldn't be a delay. The transition should be
smooth & unnoticeable. Had any of us noticed that regen braking goes away when turning
before this? No, we hadn't. That shows that the transition was working properly for
everyone in the beginning. It's only recently that these issues have begun to be reported.
Now, this loss of regen while cornering does not just happen on Ford Fusion hybrid vehicles. It
also happens on Tesla vehicles. Here is a discussion of the regen operation of a Tesla Model 3
from the Tesla Motors Club Forum. There, a contributor named AltLogic writes:45
I drove my Model 3 down my local winding mountain road. My plan was to use [only]
regenerative braking when entering the sweeping curves, and regenerative and friction
braking [together] for the sharp turns. I noticed that when I turned the steering wheel the
regenerative braking was reduced. As soon as I straightened the steering wheel, the
regenerative braking would resume at full regeneration … It was very obvious on corner
exits. While in the turn there was very little if any deceleration. As soon as the steering
wheel was close to straight there was an increase in deceleration.
Then #6DR61 responds:
I would expect this behavior. As far as I know, there is no system to detect frictional
coefficient of the road surface & tire patch. So it is prudent to reduce load from
regeneration on the drive wheels when turning with moderate lateral g forces.

Tesla Regen, Brakes and 30 R. Belt

Sudden Acceleration 1 June 2020
Then #7Skidmark adds:
Mine does it too, but I have not seen it happen on an [Model] S/X. Going around a
reasonably hard corner, regen pretty much turns itself off until the steering angle is
reduced to a certain point. It is a bit disconcerting if you’re not expecting it.
Then #8bxr140 responds:
Same thing happens on RWD [Model] S’s too. Having all your braking at the rear of the
car with high lateral load is a recipe for disaster. By turning off regen you’re forced to use
the actual [friction] brakes, which more proportionally load the car.
And #10bxr140 adds:
Yes. That’s the way the S/X models work.
Finally, Skidmark sums up the discussion:
Yes, it [the turning off of regen during a turn] is definitely put in there to reduce the
opportunity of rear-wheel traction loss leading to a spin. But, on the other hand, it also
introduces a behavior of the car that could be very unexpected to a driver. The current
behavior [turning off regen] is probably the lesser of two evils.
From these two discussions we see that all Tesla vehicles, as well as Ford Fusion hybrid vehicles,
can experience a loss of regen while cornering, This loss of regen is caused by the ABS system
turning off regen when the wheels on one side of the vehicle have a speed that differs from the
wheels on the other side of the vehicle, which happens during cornering.. It is not the result of a
design decision made by Tesla, but of the algorithms in the ABS modulator that were designed
by Tesla’s brake supplier, Bosch. And Bosch designers had no choice but to turn off regen when
their algorithms activate, because their algorithms must have control over all the braking forces
on the vehicle while the algorithms are active. And, having no control over the regen magnitude,
they must shut off regen while their algorithms are active. This is true for all electric vehicles
with regen, and not just Tesla for vehicles.
VIII. Explanation of EDR Data – The Cause of Sudden Acceleration
We are now in a position to answer the questions about the EDR data posed in Section II. If we
answer these questions in a slightly different order, it will help to better explain the events that
happened.
1. Why does regen appear to be absent in the vehicle velocity data and the longitudinal
acceleration data during the turn when the driver claims that the vehicle was in the HOLD
mode at the time and the driver was not pressing on the accelerator pedal?
Answer: Even though the Model 3 is capable of regen at a speed below the 6 mph the vehicle
was using during the right hand turn, regen appears to be absent in the EDR vehicle velocity
data and the longitudinal acceleration data because regen was turned off by the ESP hev II
module when the drive entered her turn. Turning off regen during a turn is the normal
operation of a Model 3 Tesla, and of any electric vehicle with regen, because the ABS
algorithms must know all the forces on the vehicle at the time the algorithms are time. But
the algorithms can’t know the regen forces because these forces are not controlled by the
ESP hev II module. So the ESP hev II module has no alternative but to turn off regen.
We now skip to question 5, knowing there is little doubt from the steering wheel angle data and
yaw rate data that the vehicle began to turn back to the left after initially turning to the right.
5. Why did the vehicle steer to the left as the motor speed increases, even though the driver was
making a right hand turn?

Tesla Regen, Brakes and 31 R. Belt

Sudden Acceleration 1 June 2020
 Answer: The vehicle steered to the left because the increase in vehicle speed during a right
hand turn created an oversteer situation to the right that caused the electronic stability
control (ESC) function in the ESP hev II module to activate. The function responded by
braking the outside front wheel to produce a counter-torque to reduce the oversteer to the
right. Since the outer front wheel in this case is the left front wheel, braking it caused the
vehicle to turn to the left as the motor speed increased.
We now pick up with question 2.
2. Why does the single rear drive motor speed increase at exactly the same time as the brakes
are being applied?
Answer: This is because the Engine Drag Torque Control (EDC/MSR) function has been
activated in the ESP hev II module as a result of sensing that a negative acceleration is being
applied while in a turn. This negative acceleration is clear from the longitudinal
accelerometer data. And we know that the EDC/MSR function is active at this time because
it is the correct function to apply in this case, and because other functions in the ESP hev II
module were active both before and after it became active. Therefore, it is only reasonable
that some other ECP hev II module function should be active in the intervening time. (This
is why we answered question 5 out of turn, so we could make this argument).
Now, the EDC/MSR function is intended to reduce negative acceleration caused by drag
torque produced by an engine, or in this case, to reduce regen torque caused by the drive
motor. It is not intended to reduce negative acceleration produced by friction braking.
Therefore, it checks the brake light switch (i.e., the STOP lamp switch) to determine the
cause of the negative acceleration. If the switch shows that the brake pedal is not being
pressed, then it knows that the negative acceleration is caused by regen torque, and responds
by sending a request to the drive motor to reduce the negative regen torque. This request
continues until the negative regen torque is reduced to zero. However, if the switch check
shows that the brake pedal is being pressed, then the EDC/MSR function does nothing, and
another function in the ESP hev II module reduces the negative braking torque.
What likely happened in this incident is the brake switch was faulty. As a result, when the
driver pressed on the brake pedal to cause a negative acceleration of 0.5 g’s, the brake switch
did not show that the brake pedal was being pressed. Therefore, when the EDC/MSR
function detected the negative acceleration, it checked the brake switch to find the cause and
got a false reading showing that the brake pedal was not being applied. As a result, it
concluded that the negative 0.5 g acceleration was caused by a regen torque from the drive
motor, instead of the brake system as was truly the case. This caused it to send a request to
the drive motor to reduce the negative torque by increasing the drive motor torque. Since
the drive motor torque was already zero because regen had been cut by the ESP hev II
module early in the turn, this request to increase the drive motor torque caused the drive
motor torque to increase from 0 to an RPM corresponding to 0.5 g’s of positive torque. This
is what we see in the longitudinal acceleration data, with a slight time delay of a few hundred
milliseconds caused by the command latency over the CAN bus. What has happened in this
case is that the faulty brake light switch has caused the brake pedal to behave like an
accelerator pedal. The harder the brake pedal is pressed, the more positive drive motor
torque is produced.
It is well known that brake light switches (i.e., STOP lamp switch switches) can have faults
which cause the brake lights to fail to turn on when the brake pedal is being applied. There
have been many recalls in the past for these faults, which have involved hundreds of
thousands of vehicles over the past ten years or more. Appendix 2 lists many of these recalls
by various vehicle manufacturers. From Appendix 2 it is obvious that this assumption of a

Tesla Regen, Brakes and 32 R. Belt

Sudden Acceleration 1 June 2020
 faulty brake light switch is a reasonable one that can be validated by experiment. It may also
be pointed out that, although most brake light switches are dual redundant with one of the
two switches turning ON while the other turns OFF, the EDC/MSR algorithm in this case
may only check one of the two switches to make its decision. This may increase the
likelihood for a brake switch fault that can cause a mistake in assessing the source of the
negative acceleration.
3. If the driver had her foot on the brake pedal to cause the -0.5 g longitudinal acceleration,
then how could she also have pressed on the accelerator pedal at the same time to cause the
drive motor to speed up? Tesla has stated many times in writing that when the accelerator
pedal and the brake pedal are pressed at the same time, the brake pedal always wins.
Therefore, the speeding up of the drive motor cannot be due to the driver pressing on the
accelerator pedal, and must have been caused by the vehicle itself.
Answer: From the answer to question 2, it is clear that the driver did have her foot on the
brake pedal and that pressing on the brake pedal was one of the reasons that the car drive
motor sped up to produce the sudden acceleration. The primary cause, however, for the
sudden acceleration was a fault in the brake switch that caused the EDC/MSR function to
misinterpret the negative acceleration as coming from the drive motor instead of the brake
pedal.
4. Why does the drive motor speed remain high and the vehicle continue to accelerate forward
even after the accelerator pedal reading has decreased back to zero?
Answer: The drive motor speed remains high even after the accelerator pedal reading has
decreased back to zero because the EDC/MSR function normally keeps applying the
compensating torque even after the original negative drag torque has been reduced to zero.
This is how the EDC/MSR algorithm works as discussed in Bosch patent DE10238224B4
mentioned in Section IV. This behavior is intended because the EDC/MSR function is most
often used while negotiating a curve on a highway at higher speeds. The delay allows the
vehicle to exit the curve before the positive acceleration is removed, which is good driving
practice.
6. Why does the EDR data show that the brake pedal was not pressed even though the driver
maintains that she did press the brake pedal and even though Tesla’s letter based on the
high resolution log data agrees with the driver?
Answer: The EDR data shows that the brake pedal was not pressed because the brake light
switch was faulty. The drive did actually press on the brake pedal as the driver maintains
and as Tesla agrees in a letter to the driver. This conclusion verifies the assumption made in
the answer to question 2 that the brake light switch was faulty.
7. If the driver never pressed the accelerator pedal, then why does the EDR data show that the
accelerator pedal was pressed?
Answer: Most people have 100% confidence that if the accelerator pedal data is non-zero,
then the driver was pressing on the accelerator pedal instead of the brake pedal. For them,
this is the cause of all alleged incidents of sudden acceleration. Yet, in this incident the EDR
data shows that a negative 0.5 g longitudinal acceleration was detected by an accelerometer,
which could only have been caused by the driver pressing on the brake pedal, because this
acceleration exceeds the maximum regen acceleration of a negative 0.3 g. And simultaneous
with the negative 0.5 g acceleration, the EDR data shows the accelerator pedal data
becoming non-zero, which would produce a positive longitudinal acceleration. Therefore, the
driver’s foot had to be on the brake pedal at the same time the accelerator pedal data became
non-zero. How is this possible?

Tesla Regen, Brakes and 33 R. Belt

Sudden Acceleration 1 June 2020
 To answer this question, let’s first clarify what was happening in greater detail. Figure 30
shows the sequence of events involved in this incident. The driver’s foot could either be on
the brake pedal or the accelerator pedal, and not on both at the same time. This explains the
exclusive-OR (XOR) function in Figure 30. Pressing on the brake pedal produced a negative
acceleration of 0.5 g’s which was recorded by the longitudinal accelerometer. The EDC/MSR
algorithm in the brake modulator also detected this negative acceleration, and after checking
the defective brake light switch, it interpreted the negative acceleration as a drive motor drag
torque. Therefore, it issued a request to the drive motor for a positive motor torque to offset
the negative drag torque. This request went via a CAN bus to the VSC input of the Vehicle
Torque Command Generation Function in the drive motor control system (Figure 16), where
it became a positive torque request to the drive motor torque map.

Figure 30. Sequence of events involved in this incident

Given this sequence of events, the question being asked now is: If the driver never pressed
on the accelerator pedal, then why does the EDR data show that the accelerator pedal was
pressed in this incident? If the accelerator pedal EDR data is obtained from position 1 in
Figure 30, as most people would believe, then non-zero accelerator pedal data makes no
sense because the accelerator pedal was not pressed. Therefore, it should remain zero,
causing a contradiction of the EDR data in this incident. On the other hand, if we believe
that the non-zero accelerator pedal data from position 1 is correct, leading us to believe that
the accelerator pedal was pressed, we have no explanation for how the negative 0.5 g
acceleration was produced that was recorded by the longitudinal accelerometer.
An alternative explanation for the non-zero accelerator pedal data in this incident is that the
accelerator pedal data is really obtained from position 2 in Figure 30. This would produce
non-zero accelerator pedal data either when the accelerator pedal is pressed or when a
positive torque command is issued by the EDC/MSR function. And position 2 is completely
in accordance with NHTSA’s regulation on the origin of the EDR accelerator pedal % data
because the regulation allows either the accelerator pedal position data or the throttle
position data to be recorded for the accelerator pedal % data in an internal combustion
engine. If the source of the EDR accelerator pedal data is from position 2 in this incident,

Tesla Regen, Brakes and 34 R. Belt

Sudden Acceleration 1 June 2020
 there is no contradiction of the EDR accelerator pedal data with the longitudinal
accelerometer data in this incident, and the explanation given in this paper for the sudden
acceleration remains valid.Note 6
8. Why does the EDR data show that the ABS system did not engage when both the
accelerometer data and Tesla’s letter based on the high resolution log data verify that the
ABS system did engage?
Answer: The term “ABS system” can have two meanings. In some cases, it can mean the
ESP hev II module and all its incipient functions as listed in Section IV. In other cases, it can
mean only one of the functions included in the ESP hev II module; namely the ABS function
that produces an ON/OFF modulation of the braking force to prevent wheel lock-up in the
presence of slip. It is believed that this latter meaning applies to the ABS activity indicator
logged by the EDR system. If this is the case, the EDR data indicating that the “ABS system”
failed to engage is correct, meaning that the ABS algorithm with ON/OFF modulation did
not activate. However, other functions of the ESP hev II module could still activate, and
certainly did, such as the ESC function, the EDC/MSR function, and possibly the Corner
Brake Control (CBC) function. Whether the warning light for the ABS system engaged or did
not engage is an interesting piece of information. But it is not critical to an understanding of
what happened during this event. What is more critical to know is that three ESP hev II
module functions did activate during this incident.
We have now answered all of the questions posed by the EDR data. This was done using only
one assumption; namely, that a defect in the brake light switch caused the EDC/MSR function in
the ESP hev II module to make an incorrect decision about the origin of the negative
acceleration it encountered, causing the EDC/MSR function to issue a request to the drive motor
for a large positive torque. This assumption is supported by: 1) the EDR data that shows that
the brake pedal was not pressed even though the driver maintains that she did press the brake
pedal and even though Tesla’s letter based on the high resolution log data agrees with the driver,
which proves that the brake switch was defective, and 2) Appendix 2, which gives specific
information on recalls for defective brake light switches that have affected hundreds of
thousands of passenger vehicles from nearly every manufacturer over the past ten years. If this
explanation of our original sudden acceleration incident is correct, it should also explain other
Tesla sudden acceleration incidents. Let’s now check the descriptions of other Tesla sudden
acceleration incidents to see if our explanation can apply to them also.
IX. Application to Other Tesla Sudden Acceleration Incidents
Appendix 3 lists 102 Tesla sudden acceleration incidents reported to NHTSA over the past seven
years. These incidents were obtained from a legal brief for a Class Action Suit on Tesla Sudden
Acceleration filed in a California court entitled Lee v. Tesla Inc, dated Jan 20. 2020.46 The
incidents have been divided into three categories:
A. Incidents starting while slowing down with turning to park or while accelerating
with turning to leave a parking position (70 incidents),
B. Incidents starting from a stationary position (27 incidents), and
C. Incidents starting at high speed (5 incidents).
The descriptions of the incidents omit all information about dates, locations, injuries, vehicle
damage, and follow-on communication with dealers and Tesla, leaving only information about

Note 6. Tesla Inc should be required to furnish a circuit diagram showing the origin of the EDR
accelerator pedal data. If the origin of the data is position 2 as shown in Figure 30, then this could explain
a lot of Tesla sudden acceleration incidents in which the drivers maintained that they did not press on the
accelerator pedal while Tesla cites non-zero EDR accelerator pedal data to prove that they were wrong.
Tesla Regen, Brakes and 35 R. Belt
Sudden Acceleration 1 June 2020
the incidents themselves. This was done to save the reader’s time, and to help the reader to see
similarities between the incidents.
By reading the descriptions in Category A, one can see that they are very similar to the incident
described in Section II of this paper. They involve a driver slowing down while making a turn in
preparation for either: a) parking in a perpendicular parking spot, b) driving into a driveway
prior to entering a garage, c) turning on a circular street, or d) turning a street corner. In all of
these cases, the driver can be expected to have his/her foot off the accelerator pedal and ready to
apply the brake pedal during the turn, with actually applying the brake pedal near the end of the
turn. Therefore, the explanation in this paper should apply to all of the cases in Category A,
which is about 70% of all Tesla sudden acceleration cases. The same explanation should apply
to Model S and Model X Teslas as well as Model 3 Teslas because all three models use the same
Bosch braking system and the same Tesla design for the drive motor controller.
If we read the descriptions in Categories B and C, however, we find that they differ from the
incident that produced our explanation. They involve no slowing down while turning, and
originate instead from either a stationary position (Category B) or while driving at a high speed
(Category C). Therefore, our explanation should not apply to these two categories, which
represents about 30% of all Tesla incidents. As a result, there might be more than one cause of
sudden acceleration in Tesla vehicles.
In the case of Category C incidents, one possible additional cause might be some malfunction of
the autopilot software, which could use the cruise control pathway to increase the drive motor
torque without the driver’s intervention. Some drivers have even speculated that their incidents
were caused by this explanation.
In the case of Category B incidents, it may be possible that some other braking function can
affect the drive motor torque using the same electrical pathway from the braking system to the
ESC input of the Vehicle Torque Command Generation function, to the drive motor torque map
as shown in Figure 30. For example, a self-test function in the braking system might generate a
false torque request signal to the motor torque map. Such a self-test function is likely to be
performed while the vehicle is stationary with no foot on the brake pedal. But it may also take
place with a foot on the brake pedal if a faulty brake pedal switch is indicating there is no foot on
the brake pedal.
In summary, our explanation for sudden acceleration in this incident may apply to about 70% of
all Tesla sudden acceleration incidents, which is a pretty good result. But further work is
required to explain the remaining 30%.
X. Application to Other Electric Vehicles
The first step in assessing how our explanation for sudden acceleration can apply to other
electric vehicles was to find out what braking system is used in each vehicle. This was done
because our explanation states that the cause of sudden acceleration in this case lies in the
EDC/MSR function of the brake system and how it is confused by a fault in the brake light
switch, and not in the drive motor control system. Therefore, if other electric vehicles use the
same Bosch braking system, we might expect similar sudden acceleration incidents to occur in
other electric vehicles as well.
Table 1 shows the results of our search on the braking systems of other electric vehicles. The
results are astonishing. Nearly every electric vehicle on the market today uses the Bosch
iBooster with the Bosch ESP-hev II brake module, which is exactly the same Bosch braking
system that is used by Tesla. This may come as no surprise because the Bosch iBooster allows a
lower vehicle assembly cost than a conventional vacuum modulator. And the Bosch ESP-hev II
module allows better blending of the friction brakes with the regen function under the control
of software written by the vehicle manufacturer. But Bosch has sweetened the deal even further
Tesla Regen, Brakes and 36 R. Belt
Sudden Acceleration 1 June 2020
by creating a world-wide manufacturing capability for their braking system with plants in China,
Mexico, and Germany, and by providing engineering support to vehicle manufacturers with over
2000 application engineers who can work in the manufacturers’ plants to assure successful
brake system integration. Meanwhile, other brake system suppliers like TRW/ZF have delayed
their offerings for electric vehicles due to investment problems associated with restructuring in
the world market.
Table 1. Brake systems used in electric vehicles on the market today
RWD FWD SUA ABS System &
Manufacturer Models OPD AWD
Only Only Incidents Manufacturer
BEV’s Battery EV’s
Tesla All models1 Yes Yes Yes --- Yes Bosch ESP-hev II, iBooster
BMW i35, iX3 Yes Yes --- --- Yes Bosch ESP-hev II, iBooster
Jaguar (Tata) i-Pace SUV Yes --- Yes --- Yes Bosch ESP-hev II, iBooster
Bosch SBC
Mercedes B-Class4 Yes Yes Yes --- Maybe
Bosch ESP-hev II, iBooster
VW eBKV
Volkswagen ID.3 (MEB kit) 5 Yes Yes Yes --- ---
Bosch ESP-hev II, iBooster
VW eBKV
Audi R8 e-tron No Yes Yes --- Yes
Bosch ESP-hev II, iBooster
Porsche Taycan No Yes Yes --- --- Bosch ESP-hev II, iBooster
MiniCooper Countryman ALL4 --- Yes --- --- ?
Chevrolet Bolt5 Yes --- --- Yes --- Bosch ESP-hev II, iBooster
Nissan Leaf Yes --- --- Yes --- Hitachi
Fiat Chrysler 500e Yes? --- --- Yes --- Bosch ESP-hev II, iBooster
VW eBKV
Volkswagen e-Golf (MQB kit) No --- --- Yes ---
Bosch ESP-hev II, iBooster
Hyundai Kona, Ioniq Yes --- --- Yes --- Hyundai iMEP
Honda Clarity No --- --- Yes --- Bosch ESP-hev II, iBooster
Kia Niro Yes --- --- Yes --- Bosch iBooster?
Nio ES6 No Yes No --- Bosch ESP-hev II, iBooster
Rivian R1S SUV Yes --- Bosch ESP-hev II, iBooster
ICE Hybrids With power split
Prius hybrid
Toyota --- --- Yes Yes Toyota Advics
synergy
C-MAX hybrid HEV
Ford --- --- Yes Yes TRW/ZF IBC
C-MAX Energi PHEV
ZF IBC
Chevrolet Volt --- --- Yes ---
Bosch ESP-hev II, iBooster
Cadillac CT6 PHEV --- --- Yes --- Bosch ESP-hev II, iBooster
VW eBKV w/TRW ESP
Audi A3 e-tron PHEV --- --- Yes
Bosch ESP-hev II, iBooster

This table shows two other pieces of information in addition to the braking system used,
namely: 1) whether the vehicles use rear wheel drive (RWD), front wheel drive (FWD), or both

Tesla Regen, Brakes and 37 R. Belt

Sudden Acceleration 1 June 2020
front and rear wheel drive (AWD), and 2) whether sudden acceleration incidents have been
claimed in each vehicle. The drive wheel information was included because it is believed that
the EDC/MSR function is more likely to be used by vehicles having rear wheel drive (either
RWD or AWD) than by vehicles having front wheel drive (FWD). This belief appears to be true,
because the only battery vehicles (BeV’s) that have had sudden acceleration incidents to date are
those with rear wheel drive (either RWD or AWD). Some hybrid PHEV vehicles with front
wheel drive (FWD) have had sudden acceleration incidents, but this may be the result of the ICE
engine in the PHEV causing the problem, and not the braking system.
Therefore, this table shows that all battery-powered electric vehicles (BeV’s) are susceptible to
sudden acceleration if they have: 1) a Bosch braking system, 2) rear wheel drive (either RWD or
AWD), and 3) defective brake light switches. Let’s take a look at some of the sudden
acceleration incidents that support this conclusion.
Figures 31 and 32 show a BMW i3 that has run into a wall.47 This incident took place in the
Kiamuki Shopping Center in an inner suburb of Honolulu on August 11, 2019. Although no
driver’s testimony is provided about the incident, it is clear from the photos what happened.
The driver was trying to park in a perpendicular parking space in preparation for shopping in
the Times supermarket. The parking space was near the entrance of the supermarket. As the
driver was pulling into the parking space after making a 90° turn, the vehicle’s drive motor
revved up. The BMW i3 has a single rear drive motor just like many Tesla’s. This incident is
exactly like the 70 Tesla sudden acceleration incidents of Category A in Section 9, for which our
explanation is valid.

Figure 31. BMW i3 after suddenly accelerat- Figure 32. Side view showing damage to the
ing while pulling into a perpendicular parking wall behind the parking space. Store
space.47 entrance is in the background.47

Figure 33. Store entrance near parking space. The Figure 34. Close-up of store entrance
parking space with the BMW i3 is to the left. showing the name of the supermarket.

Tesla Regen, Brakes and 38 R. Belt

Sudden Acceleration 1 June 2020
The news article mentions that this was the second such incident involving a BMW i3 in the
Honalulu area recently.
A third BMW i3 sudden acceleration case is provided by a woman named Alexandra Briseno in
California who writes:48
“I purchased a BMW i3 in my home state of California on August 21, 2016. Between
August 21, 2016 and May 7, 2017, we experienced 3 episodes of sudden acceleration. All 3
episodes occurred split seconds before the car was to make a complete stop. The 1st
incident occurred when my husband was attempting to park the car in our driveway.
The 2nd incident occurred just as I was coming to a complete stop at a traffic light.
And the 3rd incident occurred just as I was parking the car against a street curb.
The 3rd incident was the scariest because the sudden acceleration propelled my car to:
1. Collide with my tall SUV, (fortunately no visible damage because I hit the SUV's tire)
and then
2. Hit the street curb, jumped over it, and finally came to a
3. Stop just inches away from striking my husband and our small dog”.
The interesting thing about these three BMW i3 incidents is that they are exactly like the 70
Tesla sudden acceleration incidents of Category A in Section 9, for which our explanation is
valid. Yet, the BMW i3 drive motor control system uses completely different hardware and
software than the Tesla drive motor control system. This is strong evidence that the BMW i3
incidents are caused by the Bosch braking system, which is identical in the Tesla and BMW i3
vehicles.
The parallels between the BMW i3 and all Tesla vehicles run even deeper when regen is
considered. Included here are two short articles by an experienced automobile columnist who
discusses regen in the BMW i3, which the reader can compare to discussions of the Tesla in
Sections VI and VII.
“Living with BMW i3: After 2,000 Miles – Dislikes” by Tom Moloughney June 23rd, 201449
“Regen Braking is Less Aggressive. Before I start complaining, let me say that I’ve driven
just about every modern electric vehicle and plug-in-hybrid and I believe the BMW i3 has
absolutely the very best regenerative braking system on the market. Telsa probably comes in
second and the Volt, when driven in low mode, is right behind the Model S. BMW dialed back
the regen on the i3 a bit, probably in the vicinity of about 10% when compared to the ActiveE.
People who never drove the ActiveE or MINI-E won’t understand what I’m complaining about
because the i3’s regenerative braking is still strong and very smooth. It can bring the car to a
stop without using the friction brakes faster than any regenerative braking system on any other
EV will. Still, I liked it stronger like it was on the ActiveE and MINI-E. I guess regenerative
braking is like coffee. Some will prefer the Blonde Roast with cream while others want the Dark
Roast served black. Give me my regen as strong as possible please. I recommended to BMW that
they offer different regen settings and let the customer decide how strong they like it, but that
didn’t come to pass on the i3. It’s still very good, and integrates seamlessly when decelerating. I
would just prefer it a bit stronger.
When I navigate this bend in the road by my house, the regenerative braking disengages. Since
the road is also downgrade I find I have to use the friction brakes to keep from accelerating
down the hill. I didn’t have to do that in the past while driving my MINI-E or ActiveE as both
would allow the regenerative braking system to hold back the car during turns like this.
Regen Braking Disengages During Hard Turns. I’m a little surprised with the second
complaint I have with the regenerative braking. While negotiating turns, the regen sometimes
disengages which will give the sensation that the car is actually speeding up. Of course it isn’t
(unless you are going downhill), but when you are in full regen and it suddenly disengages, it

Tesla Regen, Brakes and 39 R. Belt

Sudden Acceleration 1 June 2020
does feel like the car is accelerating when in fact it just isn’t being slowed down by the
regenerative braking. During the MINI-E and ActiveE programs, I personally spoke to dozens of
people who contacted me asking if my car ever suddenly surged ahead. What was happening
with those cars was different though. If the regenerative braking system was operating and the
car hit a pothole or a bump that caused the wheels to lose traction, the traction control would
disengage the regen in an attempt to prevent the loss of control. When this happened, it would
give the driver the sensation of sudden acceleration, especially when driving downhill. This was
unsettling if you didn’t understand what was happening and typically when this happened the
owner would take the car to the dealer for service. The dealer would look it over and find
nothing wrong and give it back to them. Frustrated, many of the drivers then contacted me to
ask if anyone else had complained of this sudden acceleration problem. After explaining what
was actually happening to them they understood what was going on. I would also caution them
to always have their foot ready to press the friction brake when they were using regen to slow the
car down, especially if they were approaching the car in front of them as they were decelerating.
BMW has indeed improved the whole traction control/regenerative braking system
communication and the i3 performs much better than the MINI-E or ActiveE did when the tires
lose traction during regenerative braking. However it now disengages during cornering, and
neither of its predecessors ever did this. I can tell by how it’s working that it isn’t a flaw in my
system. It was intentionally designed to do this, perhaps to prevent the thin tires from losing
traction while negotiating hard turns. Again, it’s not a problem as long as you know it’s going to
happen and you are ready to use the friction brakes if necessary. I’ve found it mostly happens
while I’m taking a highway off-ramp that circles down under the highway overpass. It seems the
speed I’m traveling combined with the sharp, constant turn is too much and the traction control
preemptively disengages the regen in an attempt to prevent the loss of traction. I believe this is
something the dealers need to communicate to the customer. It can be a safety issue if new i3
owners aren’t prepared for it. Just like with the MINI-E and ActiveE, I’m certain there will be
customers that believe there is something wrong with their car and will take it to the dealer for
service. And just as I’m sure that will happen, I’m sure the service departments won’t have a clue
what the customers are talking about and will tell them they checked it out and car is fine.
Unless the service manager happens to read this post.”
“BMW i3 Sport One-Month Review” by Tom Moloughney 50
“Regen Braking Over Bumps and in Turns Fixed in i3 Sport. In the i3 Sport, BMW has
improved the i3’s traction control system, claiming it is now 50 times faster than before. I don’t
know how to quantify the 50 times faster claim, but I can absolutely feel the difference and it’s
definitely much better. My previous i3 would disengage the regenerative braking if I hit a bump
in the road and the car lost a little traction. Evidently, if the traction control detected any kind of
slippage it would turn off the regenerative braking, so as to not lose complete control. This
would give i3 owners the sensation that the car was actually speeding up after hitting a bump if
they weren’t accelerating or coasting when they hit the bump.
I personally had to explain what was happening to many i3 owners that reached out to me
because they thought there was something wrong with their car. Additionally, the car would
gradually decrease the level of regenerative braking in curves, as the traction control would work
to prevent slippage. While taking exit ramps at highway speeds, my car would completely cut the
regenerative braking out if the turn was sharp.
That doesn’t happen with my i3 Sport. The regenerative braking remains active when I go over
bumps and momentarily lose some traction, as well as when the car is taking curves at high
speeds.”

Tesla Regen, Brakes and 40 R. Belt

Sudden Acceleration 1 June 2020
From these articles the reader can see that regen in a BMW i3 behaves much like regen a Tesla.
We now know that this regen stopping behavior is a product of the common Bosch braking
system that shuts off regen when wheel slip is detected, and not of the different drive motor
control systems. The reader can also see from the highlighted text that the EDC/MSR function
is working properly to reduce regen in curves because the brake light switch is working normally
in this case.
XI. Application to ICE Vehicles
The Bosch ESP-hev II module used in electric vehicles is the ninth version of a modulator
module originally designed for vehicles with internal combustion engines (ICE). It was modified
for electric vehicles by changing two hydraulic valves and possibly adding an electronically
controlled plunger mechanism. But the slip control functions remained unchanged. Therefore,
these slip control functions, and specifically the EDC/MSR function, will operate the same way
in an ICE vehicle as they do in electric vehicles. It doesn’t matter if the drag force is provided by
an ICE engine or by an electric motor in regen. If the EDC/MSR function detects a drag force
while slip is occurring, it will send a request to the ICE engine to increase its torque in order to
reduce the drag torque, which will reduce the wheel slip and keep the vehicle stable. This is how
the EDC/MSR function is supposed to work. But if the brake light switch is defective, the
EDC/MSR function will confuse a brake-induced deceleration for an ICE-induced drag force,
and send the ICE engine a request to increase its torque in response to pressure on the brake
pedal. The result is that the vehicle will accelerate in response to the driver pressing on the
brake pedal, which the driver interprets as sudden acceleration. Unfortunately, when the driver
tries to explain this unusual situation to others, they invariably accuse him of “stepping on the
accelerator pedal instead of the brake pedal”. This accusation is especially effective in courts of
law when manufacturer’s attorneys tell the judge and jury that the only way the vehicle can
speed up is to step on the accelerator pedal. So the driver must have been confused.
ICE vehicles that are most susceptible to this sudden acceleration mechanism are vehicles with
rear wheel drive, which includes nearly all pickup trucks and luxury sedans, but also some sport
vehicles such as two-door coupes. In case the reader still finds it difficult to believe that this
EDC/MSR function can cause sudden acceleration in ICE vehicles, he is asked to read a paper by
the author entitled “Unintended Acceleration with a Confirmed Cause – Smaller Tires in
Front”, which is caused by this same EDC/MSR function.51
XII. Comments on Testing
In the event that a sudden acceleration incident occurs in which the driver claims that the
vehicle accelerated while his/her foot was pressing on the brake pedal, the following tests should
be performed:
1. Check the OBDII port to see if any DTC codes were set for a bad brake light switch.
2. Check the EDR summary to see if the ABS warning light was off during the incident. The
ABS function is likely to have been turned OFF by a defective brake light switch.
3. Test the brake light switch for proper electrical operation of both switches.
4. If the vehicle is still intact, operate the vehicle in the same manner it was being operated
during the incident to see if one can reproduce the sudden acceleration.
If one is trying to induce sudden acceleration on a vehicle that has not experienced sudden
acceleration by making the brake light switch defective before testing the vehicle while stopping,
be aware that some methods of generating brake switch faults may be incompatible with the
operation of the brake functions you are trying to test as a result of built-in test functions. It
may require some experimentation to produce the correct fault.

Tesla Regen, Brakes and 41 R. Belt

Sudden Acceleration 1 June 2020
 XIII. Summary and Conclusion
The EDR log data from a Tesla sudden acceleration incident has been analyzed. To explain the
EDR data, the operation of Tesla’s drive motor control system and braking system were
examined. As expected, friction braking and regen operation are completely separate with no
blending, with the possible exception of Tesla models having a single rear drive motor. The
braking system, however, also includes several vehicle stability control functions that can have a
profound effect on regen operation in the presence of wheel slip, such as stopping regen when
going over bumps and while turning corners. One of these slip control functions called
Electronic Drag Control (EDC/MSR) can even cause the drive motor to speed up if regen is
causing slip in the rear drive wheels that can lead to an oversteer or understeer. This same slip
control function can be misled by a defective brake light switch to confuse a brake-induced
deceleration for a regen-induced deceleration, in which case as the driver presses harder on the
brake pedal, a larger positive motor torque is produced. This is believed to be the cause of
sudden acceleration in the incident under consideration. To see if this same mechanism could
explain other Tesla sudden acceleration incidents, NHTSA contact reports were examined. It
was found that over 70% of all Tesla sudden acceleration incidents could be explained by this
cause. Similarities between the braking systems for Tesla and non-Tesla vehicles imply further
that this same cause can explain many sudden acceleration incidents in non-Tesla electric
vehicles having rear-wheel drive (RWD) or all-wheel drive (AWD). Finally, it is concluded that
vehicles having internal combustion engines along with rear wheel drive are susceptible to
sudden acceleration from this cause if the brake lights are defective. Comments for testing
vehicles for this mechanism are provided.
XIV. Acknowledgements
The author would like to thank the driver named ABHAR on the Tesla Motors Club forum for
her foresight and determination in obtaining the EDR data and for providing it to the
author. Few sudden acceleration victims are willing to spend the time, effort and money needed
to find the cause of their incident. But this data was essential for determining the cause of
sudden acceleration in this case. Without her initiative, this study would not have been possible.

Tesla Regen, Brakes and 42 R. Belt

Sudden Acceleration 1 June 2020
 Appendix 1. Bosch iBooster
The Bosch iBooster has several advantages over a traditional vacuum booster that make it the
booster of choice for any electric vehicle manufacturer, whether it be Tesla or any other
manufacturer of a battery or hybrid vehicle on the market. These advantages include, in
descending order of importance:
1) Electronic sensing of the brake pressure & brake pedal position. Electronic sensing
allows better blending of regen with the friction brakes for vehicles having either front
wheel drive or rear wheel drive, or with regen on the accelerator pedal or on the brake
pedal. With more electric vehicles entering the market, regen will become more widely
used, and more manufacturers will desire either the iBooster or one of its competitors.
2) Cheaper to assemble an electric vehicle because no vacuum source is required, either
from an engine manifold or an external vacuum pump. This means fewer parts at the
vehicle level and simpler assembly operations in which electrical connections replace
vacuum plumbing connections, both of which lead to lower vehicle production cost.
3) Software adjustability of booster operation by the vehicle manufacturer, including the
jump-in value and the boost value. This allows the vehicle manufacturer to tailor the
stopping performance and pedal feel for different driving scenarios even after vehicle is
manufactured. This variability is not required for good braking performance, but helps
to sell the electric booster to vehicle design engineers who can pass it along if they
desire to vehicle buyers as a driver-selectable braking feature.
4) Smaller physical size than vacuum boosters having a large vacuum diaphragm. This
allows more room under the hood for other components, which provides greater
vehicle design flexibility. .
5) Faster operation than a vacuum booster. This is a bonus feature that enables shorter
stopping distances than a vacuum booster. But vacuum boosters have not had a safety
problem with stopping distance to date, so its improvement may have little effect on
safety statistics.
6) Maintains hydraulic control by the driver in the event of booster failure. This is a
necessary safety feature that is found in vacuum boosters also. So it is not really an
advantage over vacuum boosters, but merely eliminates a possible safety concern.
The iBooster has the following disadvantages, which are considered relatively minor by most
manufacturers:
1) The iBooster is more expensive than a vacuum booster. This is because an electric
motor with its associated gearing, sensors, and electronic PID controller with output
drive transistors is more expensive than a vacuum diaphragm. But this higher cost at
the component level is offset by a lower assembly cost at the vehicle level.
2) The maximum boost of the iBooster is limited by the electric motor torque (a motor
design feature) and the motor current (a 12V battery & wiring feature). The boost of a
vacuum booster can also be limited by the diameter of its diaphragm, which is solved
by replacing it with a larger diameter diaphragm. The same problem in the iBooster is
solved by substituting a more powerful electric motor when additional torque is
needed. Bosch can supply the iBooster with four different electric motors to provide
the proper torque for any given vehicle need.
3) If the 12V battery or 12V DC/DC converter dies, then brake boost dies with it. This is
an unavoidable problem that is mitigated by the ability of the booster to maintain
hydraulic control of the brakes with the driver supplying the required braking force.
But boost will be lost if 12V power is lost, making it more difficult for the driver to stop
the vehicle. The government places a limit on the brake pedal force that the driver

Tesla Regen, Brakes and 43 R. Belt

Sudden Acceleration 1 June 2020
 must provide in order to stop safely when this occurs, so this controls the mitigation of
a 12V power loss.
Bosch sells the iBooster to vehicle manufacturers as a user-programmable unit. The iBooster
translates the brake pedal force applied by the driver into hydraulic pressure on the brake fluid
in the tandem master cylinder (TMC), which is the same hydraulic pressure provided to the
brake pads. When the brakes engage, the driver feels a return force on the brake pedal that
gives him a certain feel for how the brakes are being applied. This feel is affected by the transfer
function for how the brake pedal force is changed onto TMC pressure to the brakes, where the
TMC pressure is directly proportional to the vehicle deceleration designated by the letter z.
Figures 35 and 36 show how this transfer function can change as different parameters are
adjusted by software programming.
Figure 35 shows three different transfer functions; a) a gradual one labeled “Sport”, b) a rapid
one labeled “Comfort”, and c) an in-between one having no label. The in-between one is shown
with a hysteresis that gives it a slightly different transfer function during brake engagement than
during release. The other two transfer functions have a similar hysteresis behavior. The three
transfer curves are distinguished by having different values of the same adjustable parameters
labeled by the circled numbers in Figures 33 and 34. These parameters will now be discussed.

Figure 35. iBooster TMC pressure Figure 36. iBooster parameters software adjustable
versus pedal force showing three to vary the pedal feel. Circled numbers indicate
possible transfer curves.52 transfer curve properties shown in Figure 35.52
The cut-in parameter determines the “play” in the brake pedal, which is the amount that the
pedal must be pressed before the driver feels a return force. If this value is too high, the driver
feels that brake operation is delayed, giving an uncertain feeling that the brakes may not be
working properly. If the value is too low, then the brakes may be applied at all times, causing
undue wear on the brake pads. This parameter is not software programmable in the iBooster,
but may be adjusted by changes in the mechanical linkage between the pedal and the piston in
the master cylinder.
The jump-in parameter determines the force on the brake pads when they are initially applied,
which determines the force the driver feels on the brake pedal. If the jump-in value is too low,
the driver doesn’t feel the brakes coming on, and therefore thinks they may be defective. If the
jump-in value is too high, the driver feels that the brakes are “grabbing” and thus thinks the
brakes are too sensitive. During a parking maneuver, it is better to have a small jump-in value
to allow easier maneuvering. Sport drivers also like to have a low jump-in value because it gives
them finer modulation of the braking force during various racing maneuvers, like when entering
and exiting turns. But during panic braking, it is better to have a high jump-in value for faster
brake application. Jump-in is a mechanical feature than cannot be eliminated in conventional

Tesla Regen, Brakes and 44 R. Belt

Sudden Acceleration 1 June 2020
vacuum boosters because it is a feature determined by mechanical tolerances, making it difficult
to control within fine limits. Electric boosters allow software control of the jump-in value, and
designers usually try to adjust it to give the same feel that conventional vacuum boosters have
provided for drivers over many years.
The boost ratio is the slope of the boost curve, or the change in pressure to the brake pads with a
change in pedal force. This is not the same as the boost magnitude, or merely boost, which is
the ratio of the instantaneous brake pressure to the instantaneous pedal force. Boost ratio is the
rate at which boost is applied, which determines how rapidly the brake pressure changes with
pedal force, or how easily the brake force can be modulated. Sport drivers like to have a low
boost ratio because it gives them finer modulation of the braking force during various racing
maneuvers, like when entering and exiting turns. Most ordinary drivers, especially shorter
drivers and older drivers, prefer a high boost ratio because it makes it easier for them to stop
quickly without applying a large amount of pedal force that requires pressing one’s back against
the seat back, which they can’t do. The boost ratio of the iBooster can be customized by software
changes, which adjust how rapidly the current to the electric motor changes with changes in the
pedal force
Hysteresis is the change in pedal force needed to get the same brake pressure to the pads for an
increasing transfer function and a decreasing transfer function. Hysteresis is felt by the driver
as a slight increase in the pedal pressure as he releases the brake pedal, which is an annoying
feeling that suggests the brakes are grabbing or defective. The hysteresis should be as low as
possible for good brake operation. The iBooster minimizes hysteresis by using slightly different
amounts of electric motor torque for the increasing and decreasing transfer curves, with the
amounts being software programmable.
Runout pressure is the TMC brake pressure at which brake boost ceases, leaving only the un-
boosted transfer of brake pedal force to the brake pads. Above the runout pressure the slope of
the transfer curve becomes equal to the slope without any boost, which means that the driver
must supply a much higher brake pedal force to stop the vehicle. Heavier vehicles require
higher values of runout pressure, so runout pressure must be tailored to the vehicle weight. The
runout pressure is determined by the maximum torque that the booster motor can supply. This
maximum torque is determined either by the maximum current to the motor or the maximum
torque that the motor is designed for at some maximum current. In the first case, it is possible
to increase the motor torque by software adjustment of the motor current. In the latter case, a
more powerful electric motor can be substituted to get a higher torque. Bosch can supply the
iBooster with four different electric motors to provide different amounts of torque to meet the
needs of different vehicle weights.
Knowing how these iBooster parameters can be changed by software to change the pedal feel,
one might wonder how Tesla was able to revise their braking system so quickly in just a two
weeks to pass new braking tests by slashing 20 feet from the stopping distance after their Model
3 failed Consumer Reports tests with a braking distance of 152 feet from 60 mph53, 54. It would
appear that the changes they made affected the boost magnitude. Although a change in boost
ratio could affect how quickly the boost is applied, the test involves jamming on the brakes in an
emergency braking maneuver, so the time it takes for the boost to increase is so small that
differences in boost rate are minimal. Also, a change in the brake force distribution curve was
probably not made because regen is not a factor at such high braking decelerations and the
brake distribution curve should have already been optimized for proper front wheel braking.
Instead, they likely changed the runout pressure, i.e., the maximum boost, by either increasing
the current to the electric boost motor, or more likely, by changing the boost motor itself to one
having a higher torque output. This would mean that their initial brake system was under-
designed for the weight of the vehicle (Tesla vehicles are heavier than many ICE vehicles

Tesla Regen, Brakes and 45 R. Belt

Sudden Acceleration 1 June 2020
because of the weight of the battery) and that a higher runout pressure, or higher boost value,
was needed to give a higher braking force for the higher weight involved. This could have been
done easily with a two-week turnaround, and would have been an acceptable solution as long as
all future production vehicles use the same electric boost motor and electric current values as
the iBooster used in the tests.

Tesla Regen, Brakes and 46 R. Belt

Sudden Acceleration 1 June 2020
 Appendix 2. Recalls on Brake Switch Failures
Over the past twenty years, nearly every vehicle manufacturer has had a problem with faulty
brake light switches. Hundreds of thousands of vehicles have been recalled to replace these
faulty brake light switches. In nearly every case, the faulty switches have caused the brake lights
to fail to turn on when the brake pedal was pressed. The following are some of the cases that
have been reported.
1. Hyundai. In 2011, Hyundai announced a recall of 227,000 Hyundai vehicles for faulty brake
light switches. The recall covered eight models from August 2002 to August 2011. The
brakes would still work and the brake lights would still work, but the faulty switches could
cause a delay of up to two seconds in the brake lights turning on.55
In 2011, Hyundai announced a second recall of 99,500 2009-2011 Hyundai Accents for a
malfunctioning brake light switch that could cause the brake lights to not illuminate when
the brake pedal is depressed or could cause an inability to deactivate the cruise control by
depressing the brake pedal.56
2. Subaru. On March 2, 2019, Subaru announced a recall of 1,303,530 Subaru vehicles for
faulty brake light switches. The recall stated: “When exposed to certain contaminants, the
brake light switch may malfunction, preventing the brake lights from illuminating”.
According to a National Highway Traffic Safety Administration Safety Recall Report on this
issue, silicone gas may seep into the brake lamp switch housing. If this happens when
closure of the contacts occurs, silicon dioxide may form a deposit layer on the switch contact
terminal. The contact switch may fail to operate once this compound buildup reaches a
certain level.57
In the recall notice with Japanese regulators, Subaru said that certain chemical compounds
from everyday products, including fabric softener, cosmetics or even nail polish could cause
an insulating layer of film on the switches that could interfere with the lights being able to
turn on properly and also prevent the engine from turning on. 58
3. Kia. In 2013, Kia announced a recall of 8000 vehicles between 2008 and 2011 to fix a brake
light switch that might malfunction, meaning brake lamps could fail to operate or the cruise
control might not deactivate when the brake pedal is depressed. In Safety alert
A12/0592/13, Kia stated: “A fault in the brake light switch may result in the brake light not
coming on when the brake pedal is pressed, or in the brake light remaining on. Moreover, in
vehicles with automatic transmission it may not be possible to move the lever out of the ‘P’
position. The electronic stability control (ESC) may fail to operate, causing the ‘ESC OFF’
warning lamp on the dashboard to light up.”59
4. Suzuki. In 2019 Suzuki announced a recall for a stop lamp switch can fail in a manner that
results in failure of the stop lamp to illuminate when only the rear brake is used. … In
October and November 2018, SMC conducted an in-depth investigation of how
the structure of the stop lamp switch might contribute to the reported problems. Suzuki
found that, for the stop lamp switch produced by the Chinese supplier, the part of the
switch that moves against the switch body when the stop lamp is activated and
deactivated can become more worn than the corresponding part in the switch produced
by the Japanese supplier (the contact part of the switch produced by the Japanese
supplier is coated with grease and the contact part of the switch produced by the Chinese
supplier is not). .. In December 2018 to April 2019, Suzuki continued to conduct
reproduction testing. With repeated switch operation in a corrosive environment, Suzuki
was able to reproduce failure of the stop lamp switch.60

Tesla Regen, Brakes and 47 R. Belt

Sudden Acceleration 1 June 2020
5. Toyota. In 2013 Toyota issued a recall of 70,500 2009 Camry’s and 116,00 2009 Venza’s
because of faulty stop lamp switches. The recall stated: “During installation of the contact-
type stop lamp switch into the vehicle, silicon grease may have reached the inside of the
switch, and could cause an increase in electrical resistance. If this occurs, warning lamps
could be illuminated, the vehicle may not start, or the shift lever may not shift from the
“Park” position. In some cases, the vehicle stop lamps could become inoperative.”61
6. General Motors. In 2009, GM issued a recall of 2005-2007 Pontiac G6 vehicles for a highly
elevated incident rate of improper operation of the brake lamps. Recall number
09V036000, dated January 30, 2009, stated: “Consumers could use their brakes and
discover that the brake lamps would not illuminate when the brake pedal is applied, or the
brake lamps would illuminate when the brake pedal was not applied. Primary cause of this
brake condition was fretting corrosion (oxidation) of the connector pins on the body control
module (BCM) that connect to the brake pedal position sensor (BPPS). The corrosion
prevents the BCM from receiving the correct signal voltage from the BPPS.”
7. General Motors. In 2014, GM issued a second recall of 2,440,524 vehicles with electrical
problems in the Body Control Module (BCM). The problems could cause the brake lights to
turn on when not pressing the brake pedal or not turn on when the brake pedal is applied.
GM Technical Service Bulletin 13036A – Brake Lamp Malfunction stated: “General Motors
has decided that a defect which relates to motor vehicle safety exists in the following
vehicles:
a. 2004-2012 model year (MY) Chevrolet Malibu,
b. 2004-2007 MY Chevrolet Malibu Maxx,
c. 2005-2010 MY Pontiac G6,
d. 2007-2010 MY Saturn Aura.
On these vehicles, over time an increased resistance can develop in the Body Control
Module (BCM) connection system and result in voltage fluctuations or intermittency in the
Brake Apply Sensor (BAS) circuit that can cause service brake lamp malfunction. As a result,
the service brake lamps may illuminate when the service brakes are not being applied, or
may not illuminate when the service brakes are being applied.” … “In the affected vehicles,
increased resistance in the Body Control Module (BCM) connection may result in voltage
fluctuations in the Brake Apply Sensor (BAS) circuit. These fluctuations can cause one or
more of these conditions:
a) the brake lights to illuminate without the brake pedal being pushed;
b) the brake lights to not illuminate when the pedal is pushed;
c) difficulty disengaging the cruise control;
d) moving the gear shifter out of the 'PARK' position without pushing the brake; and
disablement of crash avoidance features such as traction control, electronic
stability control, and
e) panic braking assist features”62, 63
8. General Motors. GM may also have a problem with the actual brake light switch. Here is a
discussion found on the internet about a Chevy Tahoe Suburban:
CobraEatr
When I press the brake pedal, the brake lights will come on, but there is about a 2 to 3
second delay before they come on. … The brake lights do work (all of them), but there is a
delay when I press the brake pedal to put the car in gear. The delay is about two seconds or
so. … Problem solved. … I changed the brake light switch (P/N 15128592 - for adjustable
power pedals, $25 at the Chevy dealership) and everything is now working perfectly. … It
was just the switch that went out and was making things work sometimes, but with a delay.

Tesla Regen, Brakes and 48 R. Belt

Sudden Acceleration 1 June 2020
 I opened up the old switch and there were some burnt looking metal strips in it. So it fried at
some point.
SASIRMON87
The brake light switch has two spring-loaded contacts inside that complete the circuit and
they will get kinda burned sometimes and don’t make good contact.64
9. Ford. On November 10, 2014, Ford announced a recall of 950 2014 F-150’s for an issue with
the brake pedal position switch. If this switch is incorrectly adjusted it could result in a delay
in the brake lights lighting up or not lighting up at all when the brake pedal is depressed. 65
This recall, however, did not cover the following discussion on the Ford F150 Forum
about a Ford F10 truck and a 91 Ford F150:
mr7confused
I've noticed that my brake lights activate significantly later than my pedal starts affecting the
stopping. I can stop at a light, and keep the truck stopped with the brake lights off.
mrleno
I have a 91 F150 XLT Lariat 2WD with the same problem.66
10. BMW. The following posting by a member named “c2d” was found on the Bimmer Forum:
C2d
“I just noticed tonight my brake lights are coming on a full 4 or 5 seconds after I apply
the brake...as long as I keep my foot on the brake. … New brake light switch fixed it!67
11. Tesla. The following posting by a member named “Joelgir” was found on the Tesla Motors
Club Forum on July 13, 2018:
Joelgjr,
When backing into my garage, my brake lights weren’t coming on either in the back of the
car or on the dash. I did some testing and found out that the brake lights came on during
regen, would turn off during mechanical braking, and then would turn back on when he
brake hold activated, leaving me exposed with no brake lights whenever I was braking with
the brake pedal. … Brought it into Tesla service and they were able to reproduce it. However
the part they needed, a replacement brake pedal switch, had to be ordered so I left with the
car.68

Tesla Regen, Brakes and 49 R. Belt

Sudden Acceleration 1 June 2020
 Appendix 3. Tesla Sudden Acceleration Incidents
The Teals sudden acceleration incidents in this appendix were extracted from a legal brief for a
Class Action Suit on Tesla Sudden Acceleration filed in a California court entitled Lee v. Tesla
Inc, dated Jan 20. 2020.69 All incidents have been reported to NHTSA.
The incidents have been divided into three categories:
D. Incidents starting while slowing down with turning to park or while accelerating
with turning to leave a parking position (70 incidents),
E. Incidents starting from a stationary position (27 incidents), and
F. Incidents starting at high speed (5 incidents).
The descriptions omit all information about dates, locations, injuries, vehicle damage, and
follow-on communication with dealers and Tesla, leaving only information about the incidents
themselves. This was done to save the reader’s time, and to help the reader to see similarities
between the incidents.

A. Incidents starting while slowing down with turning to park or while

accelerating with turning to leave a parking position

1. March 30, 2018 NHTSA ID NUMBER: 11082114

I WAS DRIVING AT 10 MPH, PULLED INTO MY DRIVEWAY MADE 90 DEGREES TURN AT 5
MPH, APPLIED THE BRAKES. INSTEAD OF STOPPING, CAR SPED UP AT EXTREMELY HIGH
ACCELERATION.
2. March 6, 2017 NHTSA ID NUMBER: 10958834
WAS PARKING OUR TESLA MODEL S P85+ AND SLOWLY MOVING INTO THE PARKING
SPOT NEARING THE FINAL DISTANCE IN TO THE SPOT, WHEN THE CAR SUDDENLY
LURCHED FORWARD…
3. September 26, 2013 NHTSA ID NUMBER: 10545488
WAITING TO TURN LEFT INTO THE PARKING GARAGE. WHEN IT WAS CLEAR OF
ONCOMING TRAFFIC FOR ME TO MAKE THE LEFT TURN, I RELEASED MY FOOT OFF THE
BRAKE PEDAL AND THE CAR INSTANTLY SURGED FORWARD VERY FAST.
4. September 24, 2013 NHTSA ID NUMBER: 10545230
THE CAR WAS GOING AT ABOUT 5 MPH GOING DOWN A SHORT RESIDENTIAL
DRIVEWAY. BRAKE WAS CONSTANTLY APPLIED. THE CAR SUDDENLY ACCELERATED. …
ENGINEER FROM TESLA SAID THE RECORD SHOWED THE ACCELERATING PEDAL WAS
STEPPED ON AND IT ACCELERATED FROM 18% TO 100% IN A SPLIT SECOND. … HE ALSO
SAID THERE WAS A BUILT-IN SAFE-GUARD THAT THE ACCELERATOR COULD NOT GO
BEYOND 92%. THE STATEMENTS ARE CONTRADICTORY.
5. December 14, 2019 NHTSA ID NUMBER: 11289019
WAS SLOWLY APPROACHING OUR GARAGE DOOR WAITING FOR THE GARAGE DOOR TO
OPEN WHEN THE CAR SUDDENLY LURCHED FORWARD. … TWO MARKS ON THE GARAGE
FLOOR INDICATE THE REAR DRIVE WHEELS CONTINUED TO SPIN EVEN AFTER HITTING
THE CONCRETE WALL.

Tesla Regen, Brakes and 50 R. Belt

Sudden Acceleration 1 June 2020
6. August 24, 2015 NHTSA ID NUMBER: 10758908
I PULLED INTO A PARKING SPACE MAKING A LEFT TURN INTO THE SPOT. AS I PULLED
IN, I LET MY FOOT OFF OF THE ACCELERATOR, AND APPLIED THE BRAKE. THE CAR
LURCHED FORWARD.
7. February 15, 2016 NHTSA ID NUMBER: 10836289
CAR WAS GOING VERY SLOW AS THE DRIVER WAS TURNING TO EXIT THE COMPOUND
THROUGH A GATED ENTRANCE. THE CAR SUDDENLY ACCELERATED. … ELECTRONIC
RECORDS ACCESSED ON SEPTEMBER 15, 2015 REVEALED THE FOLLOWING
INFORMATION: THE CAR WAS ORIGINALLY BEING DRIVEN AT 4.7 KM/HR WITH THE
ACCELERATION PEDAL DEPRESSED AT 2.8%. WITHIN ONE SECOND, THE ACCELERATION
PEDAL WENT FROM BEING DEPRESSED 2.8% TO 84.8% AND THE CAR INCREASED TO 10.75
KM/HR. DURING THE FOLLOWING SECOND, THE CAR’S RECORDS SHOW THE
ACCELERATOR PEDAL TO CONTINUE TO BE COMPRESSED AT 84% WITH THE SPEED
INCREASING TO 18.35 KM/HR, BUT WITH THE BRAKE PEDAL ALSO BEING
SIMULTANEOUSLY COMPRESSED. IN THE SUBSEQUENT SECOND, THE ANTI-LOCK
BRAKES INITIATED, AND THE CAR WAS BROUGHT TO A STOP A SECOND LATER. WE
REJECT THE DEALERSHIP’S DECISION THAT THIS WAS DRIVER’S ERROR. IN SUCH A
NARROW ENVIRONMENT AS THE HOUSING COMPOUND, AND DURING A SLOW AND
CONTROLLED TURN,
8. September 29, 2014 NHTSA ID NUMBER: 10639849
WHILE PULLING INTO A PARKING SPACE, THE VEHICLE SURGED FORWARD, JUMPED THE
CURB.
9. September 29, 2014 NHTSA ID NUMBER: 10639935
I PULLED SLOWLY IN TO A PARKING SPOT & MY CAR WAS AT A STOP POSITION JUST
READY TO PUSH PARK BUTTON. WITHIN A SPLIT OF A SECOND, MY CAR (TESLA) JUMPED
THE CURB. THIS IS THE LOG FILE DATA WORD BY WORD FROM THEM. “AT THE TIME OF
THE INCIDENT THAT RESULTED IN DAMAGE TO YOUR VEHICLE, YOU INCREASED THE
ACCELERATOR PEDAL POSITION FROM 1% TO 50% IN LESS THAN ONE SECOND WITHOUT
DEPRESSING THE BRAKE PEDAL. ONE SECOND LATER, YOU INCREASED THE
ACCELERATOR PEDAL TO 100% WITHOUT DEPRESSING THE BRAKE PEDAL. ONE SECOND
LATER, YOU CONTINUED DEPRESSING THE ACCELERATOR PEDAL AT 100% WITHOUT
DEPRESSING THE PEDAL; HOWEVER, THE VEHICLE'S TRACTION CONTROL ENGAGED &
THEREFORE LIMITED THE VEHICLE'S TORQUE DESPITE THE FACT YOU WERE
DEPRESSING THE ACCELERATOR PEDAL AT 100% UNTIL YOU DEPRESSED THE BRAKE
PEDAL IN THE FOLLOWING SECOND”.
10. January 8, 2020 NHTSA ID NUMBER: 11297839
AS I DROVE MY 2015 TESLA MODEL S FORWARD AND STEERED TO THE RIGHT INTO A
PARKING SPOT, I CONSCIOUSLY EASED MY RIGHT FOOT OFF FROM THE ACCELERATOR
PEDAL, THUS LOWERING MY SPEED FROM 10-9 MPH TO ABOUT 5-6 MPH. THE BRAKE
PEDAL WAS NOT PRESSED, ALLOWING REGENERATIVE BRAKING TO SLOW DOWN THE
VEHICLE. ABOUT A SECOND INTO THE PARKING TURN, THE VEHICLE SUDDENLY JOLTED
FORWARD, ACCELERATING UNEXPECTEDLY. …. I FELT AS IF THE BRAKES FAILED TO

Tesla Regen, Brakes and 51 R. Belt

Sudden Acceleration 1 June 2020
WORK. … I CAUGHT MYSELF STILL FLOORING THE BRAKE PEDAL WITH MY RIGHT
FOOT, AND CONTINUED TO FLOOR IT UNTIL I AFTER I HAD SHIFTED IT INTO PARKING
GEAR.
11. March 11, 2018 NHTSA ID NUMBER: 11078571
WHILE PULLING INTO A PERPENDICULAR PARKING SPOT IN A PARKING GARAGE, AT ~4-5
FEET FROM A WALL, THE CAR EXPERIENCED SUDDEN ACCELERATION. … THEY HAVE
TOLD ME THAT THE CAR SPEED WAS 3.6 MPH AT WHICH TIME THE DATA INDICATES
THAT THE ACCELERATOR WAS PRESSED TO AS FAR AS 100%, RESULTING IN
ACCELERATION OF THE CAR TO 7 MPH AT THE TIME OF IMPACT, WITH THIS
ACCELERATION OCCURRING OVER A 2 SECOND TIME PERIOD. THIS INFORMATION
SEEMS INCOMPATIBLE WITH THE OTHERWISE RAPID ACCELERATION TYPICALLY
EXHIBITED BY THIS CAR WITH A FULLY DEPRESSED ACCELERATOR, AND THE SHORT
DISTANCE TRAVELED (ONLY SEVERAL FEET).
12. March 24, 2017 NHTSA ID NUMBER: 10968322
AS I APPROACHED MY PARKING SPOT, I SLOWED MY CAR BY PRESSING ON THE BRAKE
AND MADE A 90 DEGREE TURN INTO MY PARKING SPOT. THE VEHICLE WAS SLOWING
MOVING FORWARD AND WAS LESS THAN 10 FEET FROM WHERE I INTENDED TO STOP
THE VEHICLE, WHEN THE VEHICLE SUDDENLY ACCELERATED ON ITS OWN AND
CRASHED INTO THE CONCRETE WALL. … I DID NOT HAVE MY FOOT ON THE
ACCELERATOR PEDAL AT THE TIME THE VEHICLE ACCELERATED INTO THE WALL.
13. September 27, 2016 NHTSA ID NUMBER: 10910065
I DROVE THE CAR INTO A PARKING SPACE. … WHEN THE CAR ACCELERATED ON ITS
OWN (UNINTENDED ACCELERATION).
14. June 17, 2016 NHTSA ID NUMBER: 10874744
CAR RAPIDLY ACCELERATED TO MAXIMUM THROTTLE DURING PARKING MANEUVER
IN A PARKING STRUCTURE. I WAS TRAVELLING AT 3MPH. CAR ACCELERATED.
15. May 11, 2016 NHTSA ID NUMBER: 10864353
MY WIFE WENT TO PULL INTO A PARKING SPOT, SHE TOOK HER FOOT OFF THE
ACCELERATOR PEDAL AND THE CAR ACCELERATED "FULLY".
16. December 15, 2015 NHTSA ID NUMBER: 10810457
ON THE AFTERNOON OF NOVEMBER 25, 2015, I WAS DRIVING INTO A STRIP MALL
PARKING LOT. I WAS GOING TO PULL INTO ONE OF THOSE SPACES WHERE YOU CAN
PARK PERPENDICULAR TO A SIDEWALK CURB. … WHEN APPROACHING THE PARKING
SPACE, THE CAR WAS ALREADY IN REGENERATIVE BRAKING MODE, AND ACCORDING
TO TESLA’S LOGS, THE CAR SLOWED DOWN TO 3.5 MPH. SINCE THE CAR HAD ENOUGH
MOMENTUM TO ROLL INTO THE SPACE ON ITS OWN, MY FOOT WAS NOT ON THE
ACCELERATOR OR THE BRAKE PEDAL. MY FOOT WAS UP RESTING ON ITS HEEL, READY
TO TAP THE BRAKE WHEN IT GOT CLOSE ENOUGH TO THE CURB. WHILE THE CAR WAS
COASTING INTO THE SPACE THE MOTOR WAS VERY QUIET. ALL OF A SUDDEN, I HEARD
A “WHIRRING” SOUND FROM THE MOTOR. I DON’T KNOW HOW BETTER TO DESCRIBE IT,

Tesla Regen, Brakes and 52 R. Belt

Sudden Acceleration 1 June 2020
THAN TO SAY IT WAS ALMOST LIKE THE MOTOR WENT FROM A STATE OF SLUMBER TO
A FULL STATE OF AWARENESS. I BELIEVE THE MOTOR WAS “WHIRRING” LOUDLY FOR
ABOUT A SECOND BEFORE THE CAR TOOK OFF AT SUCH A FAST PACE AND WOUND UP
HITTING A BRICK WALL. … ACCORDING TO TESLA’S LOGS, THE DATA READS THAT THE
PEDAL WAS DEPRESSED DOWN TO 97% AND THE CAUSE OF THE ACCIDENT WAS DUE TO
DRIVER ERROR. I STAND FIRMLY BY MY STATEMENT THAT MY FOOT WAS NOT ON
EITHER THE ACCELERATOR OR THE BRAKE PEDAL WHEN THE CAR ACCELERATED.
DATA MAY SHOW THERE WAS PEDAL DEPRESSION BUT I DID NOT DO THE DEPRESSING.
THIS WAS DUE TO UNINTENDED ACCELERATION BY THE CAR.
17. April 29, 2016 NHTSA ID NUMBER: 10862194
THE CAR WAS BEING SLOWLY PULLED INTO A PARKING SPACE AT A BANK WHEN IT
EXPECTANTLY ACCELERATED. IT JUMPED A CURB, … HE CAR WAS BEING SLOWLY
PULLED INTO A PARKING SPACE AT A BANK WHEN IT EXPECTANTLY ACCELERATED. IT
JUMPED A CURB.
18. November 19, 2019 NHTSA ID NUMBER: 11280962
MY WIFE WAS PULLING INTO A PARKING SPOT AT HER SCHOOL AT A SLOW RATE OF
SPEED (AS SHE WAS PARKING BETWEEN 2 CARS). … WHEN SHE WAS ROUGHLY ¾ OF THE
WAY INTO THE PARKING SPOT WITH HER FOOT ON THE BRAKE, THE VEHICLE
ACCELERATED ON ITS OWN AS IF IT WAS IN PERFORMANCE MODE.
19. March 15, 2018 NHTSA ID NUMBER: 11079500
I WAS ATTEMPTING TO PARK IN FRONT OF A RETAIL BUSINESS WHEN THE CAR
SUDDENLY AND UNEXPECTEDLY ACCELERATED. … THE SUDDEN AND UNEXPECTED
ACCELERATION HAPPENED WITH NO INPUT TO THE ACCELERATOR PEDA. I WAS
LIGHTLY BRAKING, INTENT ON COMING TO A COMPLETE STOP. I HAD NEARLY
COMPLETED THE PARKING MANEUVER WHEN THE CAR RACED FORWARD,
20. January 31, 2018 NHTSA ID NUMBER: 11066047
AS MY HUSBAND WAS PULLING INTO A PARKING SPACE AND COMING TO A STOP, THE
CAR SUDDENLY ACCELERATED AT HIGH SPEED, WENT OVER THE CEMENT BUMPER FOR
THE SPACE. … WE TOOK THE CAR IN TO TESLA WHO RAN A DIAGNOSTIC. THEY ADVISED
THAT THE THROTTLE WENT SUDDENLY FROM 2% TO 97% AND THEN THE BRAKE WAS
APPLIED, BUT WHEN I ASKED FOR MORE DETAILS ABOUT THE APPROACH TO THE
PARKING SPACE, FOR A COPY OF THE REPORT THEY SAID THEY WOULD NOT PROVIDE IT
WITHOUT A SUBPOENA.
21. December 15, 2017 NHTSA ID NUMBER: 11054973
WHILE ATTEMPTING TO PARK MY TESLA S75 D IN A COSTCO SPACE, THE CAR BOLTED.
22. November 1, 2017 NHTSA ID NUMBER: 11042211
WHILE PARKING, THE VEHICLE INADVERTENTLY ACCELERATED. … THE BRAKE PEDAL
WAS DEPRESSED, BUT THE VEHICLE FAILED TO STOP. … THE FAILURE OCCURRED A
SECOND TIME WHILE PARKING IN A HOSPITAL PARKING LOT. THE VEHICLE

Tesla Regen, Brakes and 53 R. Belt

Sudden Acceleration 1 June 2020
ACCELERATED WITHOUT WARNING, JUMPED A CURB, AND COLLIDED WITH A METAL
POLE.
23. February 9, 2017 NHTSA ID NUMBER: 10953656
SHE PULLED INTO HER DRIVEWAY AT HOME AND BROUGHT THE VEHICLE TO A STOP.
WITH HER FOOT STILL ON THE BRAKE, THE VEHICLE SUDDENLY ACCELERATED ON ITS
OWN FROM A STOPPED POSITION TO SPEEDS OF BETWEEN 40-60 MPH.
24. April 4, 2018 NHTSA ID NUMBER: 11083342
WHEN I PULLED INTO THE DRIVEWAY THE SELF DRIVING FEATURES ACCELERATED THE
VEHICLE CAUSING AN ACCIDENT.
25. June 16, 2017 NHTSA ID NUMBER: 10995447
VEHICLE ACCELERATED SUDDENLY AND CRASHED INTO A TREE IN FRONT WHILE
PARKING THE CAR IN A PARKING LOT.
26. February 27, 2017 NHTSA ID NUMBER: 10957394
WHEN I WAS ABOUT TO PARK THE CAR IN THE PARKING LOT (AROUND 6 MILES PER
HOUR MAY BE) IT SUDDENLY ACCELERATED.
27. January 3, 2017 NHTSA ID NUMBER: 10939234
WHILE TURNING LEFT INTO A PARKING SPOT AT A VERY SLOW SPEED, THE CAR
SUDDENLY ACCELERATED WITH EXTREME FORCE. IT RAN OVER A CURB AND
COLLIDED WITH A TREE AND A TRUCK.
28. December 14, 2016 NHTSA ID NUMBER: 10935272
I HAD PULLED INTO A PARKING LOT, PROCEEDED TO PULL INTO A SPOT ADJACENT TO A
CINDER BLOCK BUILDING. I HAD MY FOOT LIGHTLY ON THE GAS PEDAL, THEN AS I
MADE THE TURN INTO THE SPOT, MY FOOT WAS ON THE BRAKE - THE CAR LURCHED
FORWARD AND SPED UP AND THE BRAKES DID NOT STOP IT.
29. October 12, 2016 NHTSA ID NUMBER: 10915633
WHILE PARKING THE VEHICLE, IT ACCELERATED WHILE DEPRESSING THE BRAKE
PEDAL.
30. September 26, 2016 NHTSA ID NUMBER: 10909588
I WAS GOING UP BY DRIVEWAY WAITING FOR MY GARAGE DOOR TO\ OPEN. I TOOK MY
FOOT OFF THE ACCELERATOR AND WAS SLOWING DOWN WITHOUT HITTING THE
BRAKES WAITING FOR THE GARAGE DOOR TO OPEN. THE CAR TOOK OFF THROUGH THE
GARAGE DOOR AND HIT MY HUSBANDS CAR SITTING IN THE GARAGE.
31. September 19, 2016 NHTSA ID NUMBER: 10908051
WHILE TURNING LEFT TO ENTER A VERY NARROW GARAGE ENTRANCE. … I SAW THAT I
WAS IN THE POSITION THAT I COULD CONTINUE INTO THE GARAGE AND LIGHTLY
PRESSED THE ACCELERATOR TO FINISH MY TURN INTO THE GARAGE. IT WAS AT THIS

Tesla Regen, Brakes and 54 R. Belt

Sudden Acceleration 1 June 2020
POINT THAT THE CAR ACCELERATED WITH EXTREME FORCE AND WITHIN A SECOND
SLAMMED INTO A LARGE CONCRETE POLE THAT WAS JUST INSIDE THE GARAGE TO THE
LEFT. … I DID NOT HAVE EITHER FOOT DEPRESSED ON EITHER PEDAL. … AT FIRST
TESLA TOLD US OVER THE PHONE THAT THEIR LOGS SHOW THAT THE DRIVER PRESSED
THE PEDAL 100% AND THEN TAPPED THE BRAKE BEFORE IMPACT. THIS EXPLANATION
SOUNDED PHYSICALLY IMPOSSIBLE BECAUSE THE DISTANCE COVERED WAS LESS
THAN 3 CAR LENGTHS. A MONTH LATER TESLA SENT A LETTER STATING THE DRIVER
PRESSED THE ACCELERATOR 100% UNTIL THE VEHICLE SENSED A CRASH. TESLA DID
NOT RESPOND TO OUR QUERY ABOUT WHY THEIR LOG STORY HAD CHANGED. TESLA
ALSO REFUSED TO PROVIDE DATA ABOUT ACCELERATOR/BRAKE PERCENTAGE AND
CAR SPEED.
32. August 24, 2016 NHTSA ID NUMBER: 10898260
WHILE SLOWLY PULLING INTO A PARKING SPACE… MY TESLA MODEL X SUDDENLY
ACCELERATED UNDER ITS OWN VOLITION. … AT TESLA OF TAMPA, TOLD ME
VERBALLY THE LOG FROM THE EDR SAYS THE CAR WAS TRAVELING AT 6 MPH, THEN
THE ACCELERATOR WAS ADVANCED TO OVER 50% AND THEN TO 87%. THE CAR
ACCELERATED TO 20 MPH AND ABRUPTLY STOPPED. I DENIED THIS SCENARIO AND
ASKED FOR A SUPERVISOR. TESLA’S SOUTHEAST REGIONAL MANAGER MET US AT THE
BODY SHOP. HE HANDED ME A LETTER THAT HAD DIFFERENT EDR RESULTS-VEHICLE
SPEED WAS 7 MPH, PEDAL POSITION WENT FROM 3.2% TO 15.6% TO 100% AND CAR WENT
TO 14 MPH.
33. August 4, 2016 NHTSA ID NUMBER: 10893066
WHILE ATTEMPTING TO PARK, THE VEHICLE INDEPENDENTLY ACCELERATED.
34. June 7, 2016 NHTSA ID NUMBER: 10873117
WHILE ENTERING A PARKING STALL SUDDENLY AND UNEXPECTEDLY ACCELERATED
AT HIGH SPEED ON ITS OWN.
35. July 3, 2019 NHTSA ID NUMBER: 11229124
WHILE ATTEMPTING TO PARK THE CAR IN A PARKING SPACE THE CAR ACCELERATED
ON ITS OWN.
36. March 18, 2019 NHTSA ID NUMBER: 11189710
WHILE PULLING IN THE DRIVEWAY AT APPROXIMATELY 5 MPH, THE CONTACT
ATTEMPTED TO STEP ON THE BRAKE PEDAL TO SLOW DOWN THE VEHICLE. HOWEVER,
THE VEHICLE ACCELERATED THROUGH THE GARAGE.
37. December 21, 2018 NHTSA ID NUMBER: 11162968
OUR TESLA HAS HAD TWO INCIDENTS IN THE LAST THREE MONTHS OF
UNCONTROLLABLE ACCELERATION....ONCE SLOWLY GOING INTO A PARKING PLACE
AND ONCE BACKING OUT OF OUR GARAGE.
38. July 25, 2018 NHTSA ID NUMBER: 11113560
WHILE DRIVING 5 MPH, THE VEHICLE ACCELERATED RAPIDLY WITHOUT WARNING.

Tesla Regen, Brakes and 55 R. Belt

Sudden Acceleration 1 June 2020
39. May 29, 2018 NHTSA ID NUMBER: 11098517
WHILE ATTEMPTING TO PARK THE VEHICLE AT A LOW SPEED, THE VEHICLE SUDDENLY
ACCELERATED WITH THE BRAKE PEDAL DEPRESSED.
40. January 24, 2018 NHTSA ID NUMBER: 11064628
I WAS PULLING IN TO MY OFFICE PARKING SPOT. … I LET THE CAR SLOW TO ROLL
CLOSER IN FRONT OF THE CURB. ALL OF A SUDDEN THE CAR ACCELERATED. … TESLA
SAID I WAS ON THE PEDAL FOR A SECOND LONG, I APPLIED PEDAL FROM 0-18%, AND
QUICKLY APPLIED THE BRAKE.
41. September 4, 2017 NHTSA ID NUMBER: 11021371
I WAS DRIVING DOWN THE DRIVEWAY SLOWLY AND THE CAR SUDDENLY
ACCELERATED TO A VERY HIGH SPEED.
42. September 11, 2018 NHTSA ID NUMBER: 11128789
AS I WAS PULLING INTO THE PARKING SPACE THE TESLA RAPIDLY ACCELERATED.
43. April 7, 2018 NHTSA ID NUMBER: 11083755
MY CAR HAD AN UNINTENDED ACCELERATION WHEN I WAS TRYING TO PARK IN THE
PARKING LOT OF MY OFFICE.
44. March 7, 2018 NHTSA ID NUMBER: 11076619
I WAS PARKING WHEN I FELT THE VEHICLE SUDDENLY ACCELERATE AND JUMP THE
CURB, SMASHING THROUGH THE FRONT WINDOWS/DOOR OF A BUSINESS.
45. February 16, 2018 NHTSA ID NUMBER: 11073274
WHILE ATTEMPTING TO PARK THE VEHICLE AND APPLY THE BRAKES AT 5 MPH, THE
VEHICLE PICKED UP SPEED. AS A RESULT OF THE FAILURE, THE VEHICLE CRASHED INTO
A SCHOOL BUILDING.
46. June 21, 2018 NHTSA ID NUMBER: 11102931
I WAS SLOWLY MOVING INTO A PARKING SPACE IN FRONT OF A SUPERMARKET WHEN
THE CAR SUDDENLY ACCELERATED.
47. July 1, 2019 NHTSA ID NUMBER: 11228597
I DROVE TO WORK ON MAY 2, 2019, AND WAS TRYING TO PARK IN MY ALLOTTED
PARKING STALL ON THE 3RD LEVEL OF MY OFFICE PARKING GARAGE. MY PARKING
STALL WAS ALONG THE OUTER EDGE OF THE GARAGE AND THE ONLY BARRIER TO THE
OUTSIDE WAS A SET OF THREE (ABOUT ¾ " DIAMETER) CABLES STRUNG BETWEEN
OUTSIDE COLUMNS OF THE GARAGE, BESIDES A THIN PERFORATED CORRUGATED
METAL SHEET OUTSIDE THE CABLES, ACTING AS A RAIN STOP. I WAS TRYING TO
SLOWLY TURN RIGHT FROM THE PARKING AISLE TO PARK BETWEEN TWO PICKUP
TRUCKS ON EITHER SIDE OF MY STALL. WHEN I HAD TURNED PARTWAY INTO MY
STALL, MY TESLA SUDDENLY ACCELERATED INTO MY PARKING STALL AT HIGH SPEED,
HIT THE OUTSIDE CABLES, REBOUNDED BACKWARDS AND STOPPED ABOUT 10 FEET

Tesla Regen, Brakes and 56 R. Belt

Sudden Acceleration 1 June 2020
FROM THE OUTSIDE CABLES. THE CABLES SAVED ME FROM PLUNGING 3 FLOORS DOWN
FROM THE GARAGE.
48. May 23, 2019 NHTSA ID NUMBER: 11209483
THE CAR UNINTENDED ACCELERATED WHEN I PULLED INTO GARAGE AND WAIT FOR
GARAGE DOOR OPEN.
49. March 1, 2019 NHTSA ID NUMBER: 11183545
WHILE ATTEMPTING TO MAKE A LEFT TURN FROM A PARKING SPACE, THE VEHICLE
ACCELERATED BACKWARDS IN A TIGHT CIRCLE.
50. June 6, 2018 NHTSA ID NUMBER: 11100216
WHILE ATTEMPTING TO PARK THE VEHICLE, IT INDEPENDENTLY ACCELERATED.
51. November 27, 2018 NHTSA ID NUMBER: 11154380
I WAS DRIVING MY TESLA X SLOWLY WITH AUTO PILOT OFF AND AUTO STEER OFF ON A
SURFACE STREET. I WAS TRYING TO DO A U TURN. I VERIFIED THAT THERE WAS NO
TRAFFIC, THAT THERE WAS AN AREA WHERE CARS WERE NOT PARKED ON THE LEFT
SIDE OF THE STREET AND COULD USE A PARKING DRIVEWAY ON THE RIGHT THAT WILL
ALLOW MORE SPACE TO PERFORM THE U TURN. I GENTLY PRESSED THE BRAKE TO
FURTHER SLOWDOWN AND THEN VEERED A LITTLE TO THE RIGHT INTO THE DRIVEWAY
AT LESS THAN 5 MPH. I STARTED TO TURN TO THE LEFT WHILE STILL GENTLY PRESSING
THE BRAKE. AS SOON AS THE FRONT RIGHT WHEEL LEFT THE DRIVEWAY AND BEFORE I
HAD COMPLETED THE U TURN, THE TESLA ALL OF A SUDDEN, ON ITS OWN
SPONTANEOUSLY ACCELERATED TO FULL POWER (A TESLA X CAN GO FROM 0-60 MILES
IN APPROX. 4 SECONDS). MY FOOT WAS STILL ON THE BRAKE, I DID NOT MOVE IT AND
THEREFORE I DID NOT PRESS THE ACCELERATOR. THERE WERE NO OBJECTS
OBSTRUCTING THE PEDALS. IN A SMALL FRACTION OF A SECOND, THE TESLA
TRAVELED AT FULL SPEED TOWARDS THE TACOMA CAR PARKED AHEAD ON THE LEFT
(LESS THAN 20 FEET AWAY).
52. October 23, 2018 NHTSA ID NUMBER: 11142282
WHILE IN A PARKING LOT AT LOW SPEED AND COMING AROUND A TURN TO PARK,
TESLA SUDDENLY ACCELERATED FROM LOW SPEED TO HIGH SPEED.
53. July 15, 2018 NHTSA ID NUMBER: 11111431
WHEN SLOWLY PULLING INTO A PARKING SPOT, THE CAR ACCELERATED ON ITS OWN
JUMPING THE CURB. … THIS IS THE SECOND INCIDENT OF THE CAR ACCELERATING ON
ITS OWN. THE FIRST TIME THIS OCCURRED … I WAS SLOWLY DRIVING AROUND A
TRAFFIC CIRCLE IN A RESIDENTIAL AREA AND THE CAR TOOK OFF ON ITS OWN.
54. May 8, 2018 NHTSA ID NUMBER: 11092528
WHILE PULLING INTO A PARKING SPACE, THE BRAKE PEDAL WAS DEPRESSED BUT THE
VEHICLE INADVERTENTLY ACCELERATED OVER A CURB.
55. March 1, 2019 NHTSA ID NUMBER: 11183334

Tesla Regen, Brakes and 57 R. Belt

Sudden Acceleration 1 June 2020
WAS PARKING IN MY DRIVEWAY. THE VEHICLE SUDDENLY ACCELERATED, HIT THE
CLOSED GARAGE DOOR.
56. NHTSA ID Number: 11133222
TO PULL FORWARD OUT OF A PARKING SPACE (I BACKED IN), I GENTLY APPLIED THE
ACCELERATOR PEDAL TO EASE FORWARD (AND TURN LEFT) OUT OF THE PARKING
SPACE TO HEAD DOWN THE PARKING AISLE, THE CAR MOVED FORWARD MUCH FASTER
THAN EXPECTED SO I LET OFF THE PEDAL. THE VEHICLE STOPS ACCELERATING AFTER
LETTING OFF THE PEDAL. I TRY PRESSING THE PEDAL GENTLY AGAIN AND THE CAR
LURCHES FORWARD PRETTY QUICKLY AS IF I DEPRESSED THE PEDAL ABOUT HALFWAY
DOWN (I DID NOT). LETTING GO OF THE ACCELERATOR PEDAL DIDN'T SEEM TO KICK IN
REGENERATIVE BRAKING SO I PRESSED ON THE BRAKE TO SLOW THE CAR DOWN TO A
SPEED APPROPRIATE FOR A PARKING LOT…. IT'S ALMOST AS IF THE "THROTTLE MAP"
WAS MESSED UP FOR A FEW SECONDS ASKING THE CAR TO DELIVER MUCH HIGHER
CURRENT THAN IT SHOULD HAVE AT A 5-10% PEDAL POSITION.
57. August 14, 2018 NHTSA ID NUMBER: 11119991
I RELEASED THE ACCELERATOR PEDAL AND TURNED TO APPROACH THE CURB. WHEN
THE VEHICLE WAS NEXT TO THE CURB, IT SUDDENLY ACCELERATED.
58. May 22, 2018 NHTSA ID NUMBER: 11097159
DRIVER FELT SUDDEN UNINTENDED ACCELERATION WHILE PARKING THE CAR. THE
ACCELERATION HAPPENED WITHIN THE LAST COUPLE OF SECONDS WHEN THE CAR
WAS TURNING 90° AT PRETTY NORMAL SPEEDS TO PARK.
59. May 10, 2018 NHTSA ID NUMBER: 11092830
WHILE ATTEMPTING TO PARK THE VEHICLE, THE BRAKE PEDAL WAS DEPRESSED, BUT
THE WHILE ATTEMPTING TO PARK THE VEHICLE, THE BRAKE PEDAL WAS DEPRESSED,
60. May 4, 2018 NHTSA ID NUMBER: 11091970
WHILE TRYING TO PARK THE VEHICLE IN A PARKING LOT OF A GROCERY STORE. THE
VEHICLE WAS IN THE PARKING SPACE POSITION WHEN IT SUDDENLY ACCELERATED
WITHOUT ANY INPUT FROM THE DRIVER. THE BRAKES WERE APPLIED HOWEVER THEY
FELT INOPERATIVE.
61. October 14, 2019 NHTSA ID NUMBER: 11268280
I WAS TRYING TO PARK MY VEHICLE IN A PARKING SPOT I WAS TRYING TO PARK MY
VEHICLE IN A PARKING SPOT.
62. May 22, 2019 NHTSA ID NUMBER: 11209238
MY WIFE) DROVE UP MULTILAVEL PARKING LOT. … SHE INITIALLY CHOSE A HEAD IN
PARKING SPOT, THEN ATTEMPTED TO MOVE THE CAR FORWARD TO THE NEXT AISLE,
SO THE CAR WOULD BE FACING FRONT WHEN SHE GOT OUT. THE CAR PICKED UP
UNCONTROLLABLE AND EXTREMELY HIGH SPEED.
63. April 16, 2019 NHTSA ID NUMBER: 11196764

Tesla Regen, Brakes and 58 R. Belt

Sudden Acceleration 1 June 2020
VEHICLE ENTERED A SHOPPING PLAZA AT SPEED UNDER 10MPH AND SUDDENLY
ACCELERATED TO HIGH SPEED ON ITS OWN. THE BRAKES WOULD NOT WORK.
64. September 28, 2018 NHTSA ID NUMBER: 11132177
CAR WAS IN PARKING LOT (BOTH ON AUGUST12 2018 AND ON SEPTEMBER 26 201 AND
COMING TO A STOP… MY FOOT WAS NOT ON THE ACCELERATOR BUT CAR JOLTED
FORWARD.
65. September 2, 2018 NHTSA ID NUMBER: 11124067
I WAS PULLING INTO A DRIVEWAY WHEN MY FOUR WEEK OLD TESLA MODEL 3,
UNEXPECTEDLY ACCELERATED.
66. October 21, 2019 NHTSA ID NUMBER: 11269912
WAS PULLING INTO THE DRIVEWAY AT 4 MPH, THE VEHICLE ACCELERATED AND
CRASHED INTO THE GARAGE DOOR. … HE STEERING WHEEL AND BRAKES SEIZED.
67. December 30, 2019 NHTSA ID NUMBER: 11292014
I AM SLOWLY DRIVING AT THE PARKING LOT. … I AM SLOWLY PARKING MY CAR AND
THEN THERE’S A SUDDEN UNINTENDED ACCELERATION ON MY CAR WHICH I CANNOT
EVEN CONTROL OR STOP. THE ACCELERATION SPEEDS UP AUTOMATICALLY ON ITS
OWN EVEN IF I’M STEPPING ON THE BRAKE.
68. July 15, 2019 NHTSA ID NUMBER: 11231846
MY TESLA MODEL 3 JUST SUDDENLY ACCELERATED AS I WAS TURNING RIGHT TRYING
TO THE LOT TO PARK THE CAR.
69. November 4, 2019 NHTSA ID NUMBER: 11278152
I TURNED INTO A PERPENDICULAR PARKING SPACE IN A HOME DEPOT PARKING LOT. I
LET UP ON THE ACCELERATOR AS THE CAR INCHED CLOSER TO THE CURB. AS I BEGAN
TO PRESS THE BRAKE TO BRING THE CAR TO A COMPLETE STOP, THE CAR SUDDENLY
LURCHED FORWARD.
70. January 14, 2020 NHTSA ID NUMBER: 11299347
WHILE PULLING INTO THE PARKING LOT OF A SHOPPING CENTER WITHOUT WARNING
THE VEHICLE ACCELERATED.

B. Incidents starting from a stationary position

1. March 26, 2018 NHTSA ID NUMBER: 11081382

WHILE STATIONARY WITH THE BRAKE PEDAL DEPRESSED, THE VEHICLE
INDEPENDENTLY ACCELERATED RAPIDLY WITHOUT WARNING.
2. August 18, 2015 NHTSA ID NUMBER: 10749575

Tesla Regen, Brakes and 59 R. Belt

Sudden Acceleration 1 June 2020
THE VEHICLE SUDDENLY ACCELERATED WITHOUT WARNING TO ITS MAXIMUM
ACCELERATION RATE.
3. October 9, 2018 NHTSA ID NUMBER: 11139174
WHILE ATTEMPTING TO PARK THE VEHICLE IN THE DRIVEWAY, IT SPONTANEOUSLY
ACCELERATED.
4. May 1, 2017 NHTSA ID NUMBER: 10982961
AS I WAS PARKING MY CAR I HEARD AND FELT THE CAR BEGIN TO ACCELERATE. I PUT
THE CAR IN PARK AND STEP FIRMLY ON THE BRAKE PEDAL. THEN OF IT OWN VOLITION
IT MOUNTED THE ADJOINING EMBANKMENT.
5. March 8, 2016 NHTSA ID NUMBER: 10845619
I WAS STOPPED AT A RED LIGHT AT AN INTERSECTION THE EVENING OF MARCH 2, 2016,
WITHOUT ANY NOTICE, THE CAR ABRUPTLY ACCELERATED. MY FOOT WAS ONLY ON
THE BRAKE.
6. December 27, 2019 NHTSA ID NUMBER: 11291423
2015 MODEL S 85D WAS REVERSED ONTO DRIVEWAY THEN PLACED IN PARK AND DOORS
WERE CLOSED AND LOCKED. A FEW MOMENTS LATER THE VEHICLE STARTED
ACCELERATING FORWARD TOWARDS THE STREET AND CRASHED INTO A PARKED CAR.
FRONT WHEELS WERE RECEIVING POWER WHILE REAR WHEELS WHERE LOCKED AND
DRAGGING RATHER THAN WHEELS SPINNING. I REVERSED VEHICLE BACK ONTO
DRIVEWAY AND IT HAPPENED ANOTHER 2 TIMES AFTER FIRST INCIDENT WITHIN A 30
MINTUE TIME SPAN.
7. January 20, 2019 NHTSA ID NUMBER: 11171052
I HAD PLACED THE CAR IN PARK AND WAS PREPARING TO EXIT THE VEHICLE WHEN IT
SUDDENLY ACCELERATED AND HIT THE FRONT OF MY GARAGE WALL. THE CAR WAS IN
PARK AND NOT MOVING. I HAD MY HAND ON THE DOOR HANDLE TO EXIT. I
IMMEDIATELY SLAMMED THE BRAKES BUT THE CAR SEEMED TO KEEP GOING. IT JUST
SUDDENLY ACCELERATED WHILE IN PARK.
8. April 19, 2017 NHTSA ID NUMBER: 10979378
WHEN STOPPED AT THE SIGNAL, THE VEHICLE ACCELERATED INDEPENDENTLY
WITHOUT WARNING.
9. September 17, 2015 NHTSA ID NUMBER: 10764853
CAR WAS PARKED IN THE PARKING LOT. WHILE BACKING THE CAR, IT SPIN FAST OUT
OF CONTROL EVEN THOUGH BACKING WAS STARTED AT A VERY SLOW SPEED.
10. August 8, 2018 NHTSA ID NUMBER: 11118541
WHILE IN REVERSE BACKING UP TO A CHARGE STATION IN A COVERED PARKING LOT,
THE CAR WENT INTO SUDDEN ACCELERATION AND FAILED TO STOP.
11. June 9, 2018 NHTSA ID NUMBER: 11100721

Tesla Regen, Brakes and 60 R. Belt

Sudden Acceleration 1 June 2020
I WAS DRIVING UP MY DRIVEWAY. UPON REACHING THE TOP OF THE DRIVEWAY, THE
CAR SUDDENLY ACCELERATED BY ITSELF. … I HAVE 4 PASSENGERS THAT CAN ATTEST
TO THIS.
12. June 19, 2017 NHTSA ID NUMBER: 11000077
WHILE ATTEMPTING TO EXIT THE PARKED VEHICLE, IT INDEPENDENTLY
ACCELERATED. … WHEN THE BRAKES WERE APPLIED, THE VEHICLE CAME TO A STOP.
13. February 6, 2017 NHTSA ID NUMBER: 10949955
I WAS STOPPED AT A STOP LIGHT … AS THE LIGHT TURNED GREEN, I SLOWLY PRESSED
ON THE GAS TO MOVE FORWARD AND THE CAR TOOK OFF AT TOP SPEED. I HIT THE
BRAKE BUT THE CAR DID NOT RESPOND – IT DID NOT SLOW DOWN OR STOP,
14. May 10, 2016 NHTSA ID NUMBER: 10864163
INCIDENT 1: MY WIFE WAS AT A STOP SIGN. SHE REMOVED HER FOOT FROM THE BRAKE
AND BEFORE APPLYING THE ACCELERATOR THE CAR SURGED FORWARD
AGGRESSIVELY. SINCE HER FOOT NEVER TOUCHED THE ACCELERATOR SHE WAS ABLE
TO APPLY THE BRAKE AND STOP WITHIN 8-10 FEET. … INCIDENT 2: ABOUT 2 WEEKS
LATER, MAY 6, I WAS PULLING INTO MY GARAGE WITH MY WIFE AND BABY IN THE
VEHICLE. 2-3 FEET FROM THE GARAGE WALL (IN CREEP MODE) I GENTLY TOUCHED THE
BRAKE TO COME TO A STOP. AT THAT POINT THE CAR SURGED FORWARD VERY
AGGRESSIVELY. I IMMEDIATELY APPLIED HEAVY BRAKE AND WAS ABLE TO STOP THE
CAR IN A FEW FEET (SINCE MY FOOT WAS ALREADY OVER THE BRAKE PEDAL … LOGS
WERE DOWLOADED AND SHOW THAT THE ACCELERATOR WAS DEPRESSED TO 97%
POWER IN LESS THAN A SECOND AND THAT IT WAS DRIVER ERROR. … MY WIFE IS 37,
HEALTHY AND A PHYSICIAN, I AM A HEALTHY 42 YEAR OLD PILOT.
15. May 17, 2018 NHTSA ID NUMBER: 11096621
I NEEDED TO BACK INTO A SPOT TO BE ABLE TO TAKE PART IN THE FUNERAL
PROCESSION. … AT THE SAME MOMENT THAT I STARTED TO BACK IN, THE CAR SURGED
AND MADE AN ARC IN THE PARKING LOT ONLY STOPPING AFTER I HIT A PARKED CAR.
… I WAS TOLD IT COULD BE THE AUTO PILOT SYSTEM THAT COULD BE A FAULT EVEN
THOUGH I HAVE NEVER USED IT.
16. April 4, 2018 NHTSA ID NUMBER: 11083342
WHEN I PULLED INTO THE DRIVEWAY THE SELF DRIVING FEATURES ACCELERATED THE
VEHICLE CAUSING AN ACCIDENT.
17. August 24, 2016 NHTSA ID NUMBER: 10898260
MY WIFE WAS WAITING IN LINE AT A CAR WASH. WHILE WAITING FOR THE CAR IN
FRONT OF HER TO MOVE OUT OF THE CAR WASH TUNNEL, SHE HAD HER FOOT FULLY
PRESSED ON THE BRAKE PEDAL AND THE “HOLD” SIGN WAS SHOWING ON THE
INSTRUMENT PANEL SCREEN. APPROXIMATELY 2 MINUTES LATER, SHE RELEASED THE
BRAKE PEDA. BUT BEFORE SHE COULD PRESS THE GAS PEDAL, THE CAR SUDDENLY
MOVED FORWARD BY ITSELF AT ALMOST FULL ACCELERATION. MY WIFE
IMMEDIATELY ATTEMPTED TO STOP THE CAR BY PRESSING THE BRAKE PEDAL IN FULL.
THE CAR SLOWED DOWN, BUT DID NOT FULLY STOP AND KEPT MOVING FORWARD
UNTIL IT HIT THE CAR IN FRONT OF IT.

Tesla Regen, Brakes and 61 R. Belt

Sudden Acceleration 1 June 2020
18. July 7, 2017 NHTSA ID NUMBER: 11003716
MY TESLA MODEL X 2016 ACCELERATED ON ITS OWN AFTER I COME TO A COMPLETE
STOP AT A PARKING LOT.
19. January 28, 2018 NHTSA ID NUMBER: 11065563
AT OUR DRIVEWAY, COMING BACK FROM SHOPPING, SUDDENLY THE CAR TOOK OFF BY
ITSELF. IT ACCELERATED TREMENDOUSLY WITH THE SPEED WHICH WAS FELT AROUND
65-80 M/HR, HITTING OUR HEAVY GARAGE DOOR, WHICH WAS CLOSED AT THE TIME.
20. November 23, 2017 NHTSA ID NUMBER: 11048161
WE WERE IN A PARKING LOT. … OUR CAR WAS IN STATIONARY POSITION. WE WERE
TRYING TO FIGURE OUT WHERE TO PARK. DURING THIS TIME, OUR MODEL X, TOOK OFF
WITH UNINTENDED ACCELERATION AND WE COULD NOT CONTROL THE CAR. IT WAS
INSTANTANEOUS.
21. December 3, 2018 NHTSA ID NUMBER: 11155579
CAR WAS STOPPED PRIOR TO ENTERING CAR GARAGE. CAR SUDDENLY REVERSED THEN
SPED UP AUTONOMOUSLY AND CRASHED INTO THE HOME.
22. June 18, 2018 NHTSA ID NUMBER: 11102347
WHILE APPROACHING A RED LIGHT, THE SPEED INCREASED AND THE VEHICLE FAILED
TO RESPOND WHEN THE BRAKE PEDAL WAS DEPRESSED. THE VEHICLE DROVE
THROUGH THE TRAFFIC LIGHT AND CRASHED.
23. December 15, 2019 NHTSA ID NUMBER: 11289172
USING THE SELF-PARK SYSTEM, FROM A STATIONARY POSITION AND PRESSING THE
TOUCH SCREEN PROMPTS TO USE THE SELF-PARK, AND HAVING HANDS AND FEET
AWAY FROM THE GAS, BRAKE AND STEERING WHEEL, AND PARKING BETWEEN TWO
CARS, THE CAR ACCELERATED BACKWARD ON ITS OWN AND I HAD TO SLAM MY FOOT
ON THE BRAKE TO STOP IT BUT IT STILL HIT A POLE GOING IN REVERSE.
24. November 26, 2018 NHTSA ID NUMBER: 11154132
TESLA MODEL 3 (2018) PROCEEDED TO ACCELERATE WHILE BRAKE PEDAL WAS
DEPRESSED GOING THROUGH AN INTERSECTION. … THE VEHICLE FAILED TO RESPOND
TO PEDAL INPUT. … AUTOPILOT WAS ENGAGED PRIOR TO CRASH.
25. November 30, 2019 NHTSA ID NUMBER: 11282993
WE WERE DRIVING OUR 5-MONTH-OLD TESLA MODEL 3. WE WERE COMING TO A STOP
AT A STOPLIGHT IN HEAVY TRAFFIC, WHEN SUDDENLY THE CAR ACCELERATED ON ITS
OWN. MY HUSBAND HAD TO BRAKE SUDDENLY TO PREVENT RUNNING INTO ANOTHER
CAR.
26. October 8, 2019 NHTSA ID NUMBER: 11267131
MY WIFE WAS TRYING TO PARK THE CAR IN OUR GARAGE WHEN THE CAR SUDDENLY
SPED UP AND HIT OUR INNER GARAGE WALL. IT THEN REVERSED BACK AND HIT THE
WALL A SECOND TIME. THEN IT AUTOMATICALLY REVERSED AND PARKED ON THE
NEIGHBOR’S DRIVEWAY.

Tesla Regen, Brakes and 62 R. Belt

Sudden Acceleration 1 June 2020
27. May 10, 2019 NHTSA ID NUMBER: 11206931
I WAS STOPPED AT A RED LIGHT, SECOND IN LINE. MY CAR LUNGED BACKWARDS
HITTING THE CAR BEHIND ME. I HAD DONE NOTHING TO MOVE THE VEHICLE.

C. Incidents starting at high speed

1. November 5, 2019 NHTSA ID NUMBER: 11278322
WHILE DRIVING 65-70 MPH FOR THIRTY MINUTES WITH THE AUTO PILOT ENGAGED, THE
VEHICLE SPONTANEOUSLY ACCELERATED WITHOUT WARNING.
2. June 15, 2017 NHTSA ID NUMBER: 10995382
GOING THE SPEED LIMIT, & RELEASED THE ACCELERATOR SO THAT "ENGINE BRAKING"
WOULD SLOW THE CAR. BUT I WAS SLOWING DOWN FASTER THAN NECESSARY WITH
~100 YARDS REMAINING TO REACH THE INTERSECTION. I LIGHTLY PRESSED THE
ACCELERATOR TO STOP ENGINE BRAKING BRIEFLY, THEN RELEASED THE
ACCELERATOR TO REINSTATE "ENGINE BRAKING". INSTEAD OF SLOWING, MY CAR
BEGAN TO WILDLY ACCELERATE. … ANOTHER PIECE OF INFORMATION: 2 TIMES WITHIN
THE PRIOR WEEK, WHEN GETTING INTO MY CAR (AWAY FROM MY HOME), THE CAR
INDICATED "PREPARING TO DRIVE" OR SOMETHING TO THAT EFFECT (NOT A MESSAGE I
NORMALLY SEE), THE MESSAGE PERSISTED, SO BOTH TIMES I HAD TO REBOOT THE
COMPUTER SYSTEM, TO MAKE THE CAR DRIVE-ABLE. COULD THIS BE AT ALL RELATED
TO THE ACCIDENT ON JUNE 9TH?
3. August 7, 2018 NHTSA ID NUMBER: 11118315
THE CAR AUTO ACCELERATED ON A SURFACE STREET AND CRASHED.
4. November 30, 2019 NHTSA ID NUMBER: 11282996
ON 11-27-19, OUT TESLA MODEL 3 WHICH IS 5 MONTHS OLD, SUDDENLY ACCELERATED
IN TRAFFIC. MY HUSBAND WAS DRIVING ABOUT 20 MILES AN HOUR AT THE TIME. THIS
ALSO OCCURRED ON 11-29-19,
5. October 1, 2019 NHTSA ID NUMBER: 11265452
AS VEHICLE WAS TRAVELING ON THE HIGHWAY, THE DRIVER REPORTEDLY TRIED TO
SLOW THE VEHICLE AS IT APPROACHED STOPPED/SLOWED TRAFFIC. VEHICLE DID NOT
RESPOND AND DRIVER WAS NOT ABLE TO SLOW IT BEFORE IT CRASHED INTO STOPPED
TRAFFIC.

Tesla Regen, Brakes and 63 R. Belt

Sudden Acceleration 1 June 2020
 References

1 The Tesla Team, “There is no “unintended acceleration” in Tesla vehicles”, January 20, 2020
https://www.tesla.com/es_ES/blog/no-unintended-acceleration-tesla-vehicles?redirect=no
2 Steve Hanley, “Tesla Responds to “Sudden Unintended Acceleration” Petition to NHTSA”, Jan

21, 2020. https://cleantechnica.com/2020/01/21/tesla-responds-to-sudden-unintended-

acceleration-petition-to-nhtsa/
3 Nathaniel Michaluk, “One Pedal Driving – Rapid Feature Development with Simulink”,

Mathworks Automotive Conference, May 9, 2017.

4 J.J. Van Boekel,I.J.M Besselink, and H. Nijmeier, “Design and Realization of a One-Pedal

Driving Algorithm for the TU/e Lupo EL”, World Electric Vehicle Journal, Vol 7, Kintex, Korea,
May 3-6, 2015.
5 J. Wang, “Battery Electric Vehicle Energy Consumption Testing and Prediction: A Practical

Case Study”, PhD Thesis, Eindhoven Univeristy of Technology, Eindhoven, the Netherlands,
2016
6 Dongbin Lu, Minggao Ouyang, Jing Gu, and Jianqiu Li, “Instantaneous Optimal Regenerative

Braking Control for a Permanent Magnet Synchronous Motor in a Four-Wheel Drive Electric
Vehicle”, Proceedings of the Institution of Mechanical Engineers, Part D; Journal of Automobile
Engineering, published online February 2014.
7 Xiaofeng Ding, Jiawei Cheng, and Feida Chen, “Impact of Silicon Carbide Devices on the

Powertrain Systems in Electric Vehicles”, in Energies 2017, edited by Rui Xiong, 10(4), p533.
https://doi.org/10.3390/en10040533, found at https://www.mdpi.com/1996-1073/10/4/533
8 EPA Vehicle Chassis Dynamometer Driving Schedules (DDS), Vehicle and Fuel Emissions

Testing, United States Environmental Protection Agency, https://www.epa.gov/vehicle-and-

fuel-emissions-testing/dynamometer-drive-schedules
9 Wugz, “Model 3 AWD tested from 100-0 km/h – ROLL Mode vs. HOLD Mode”, Reddit,

https://www.reddit.com/r/teslamotors/comments/dv671z/model_3_awd_tested_from_1000_
kmh_roll_mode_vs/
10 Ibid, Note 7.
11 Rayad Kubiasi, “Adaptive Regenerative Braking in Electric Vehicles”, Dissertation for Doctors

der Ingenieur wissenchaften (Dr. –Ing), Karlsruher Institut fur Technologie (KIT), December 1,
2018.
12 Selim Oleksowicz, Mariusz Ruta, Keith Brunham, Eddie Curry, and Hector Garces, “Legal,

Safety, and Practical Regenerative Braking Control Challenges”, Measurement and Control,
Vol 46, No. 9, November 2013.
13 This figure is found in the following ten US Tesla patents and patent applications:

US7739005B1, US7742852, US7747363, US9162586B2, US2014/0257613A1,

US2016/0009197A1, US8453770B2, US8761985B2, US2010/0222953, US2013/0241445. All
have incorrect figures showing an addition operation instead of a modulation operation. The
last four patents show a VSC input to the Vehicle Torque Command Generation Function.
14 Ibid, Note 10.
15 Bosch iBooster, EN | IAA 2013, https://www.youtube.com/watch?v=f8RsWGs7Ls0&feature

=youtu.be
16 Robert Bosch GmbH, “iBooster – Vacuum Independent Electro-Mechanical Brake Booster”,

product spec sheet 292000P12J-C/CCA-201308-En, 2013.

17 https://www.bosch-mobility-solutions.com/en/products-and-services/passenger-cars-and-

light-commercial-vehicles/driving-safety-systems/brake-booster/ibooster/
18 Ibid, Note 14.
19 https://www.bosch-mobility-solutions.com/en/products-and-services/passenger-cars-and-

light-commercial-vehicles/driving-safety-systems/regenerative-braking-systems/esp-hev/
20
http://www.myshared.ru/slide/885225/

Tesla Regen, Brakes and 64 R. Belt

Sudden Acceleration 1 June 2020
21 US patent 9315182B2, “Braking System”, by Ryan A. Kuhlman, assigned to Robert Bosch
GmbH, April 19, 2016.
22 US patent 9586486, “Control Unit for a Recuperative Brake System of a Vehicle and Method

for Braking a Vehicle”, by Michael Kunz and Stefan Strengert, assigned to Robert Bosch GmbH,
Mar 7, 2017.
23 Max Boerboom, ”Electric Vehicle Blended Braking Maximizing Energy Recovery While

Maintaining Vehicle Stability and Manueverability”, Master’s Thesis 2012:01, Chalmers

University of Technology, Goteborg, Sweden, Jan 2012.
24 US patent 6535809B1, “Vehicle Engine Torque Control with Engine Drag Control Mode”, by

Keith Beyer, Eric, Krueger, Vincent Marchese, Todd Brown, and Alan Lustre, assigned to Delphi
Techlologies, Inc; March 18, 2003.
25 German patent DE10238224B4. “Curve-Dependent Engine Drag Torque Control”, by

Thomas Sauter, assigned to Robert Bosch GmbH, October 16, 2003.

26 US patent 6535809B1, “Vehicle Engine Torque Control with Engine Drag Control Mode”, by

Keith Beyer, Eric, Krueger, Vincent Marchese, Todd Brown, and Alan Lustre, assigned to Delphi
Techlologies, Inc; March 18, 2003.
27 US patent application 2003/0221889A1, “Curve-Dependent Engine Drag Torque Control”,

Thomas Sauter, March 27, 2003.

28 US patent 7681960B2, “Method and System for Improving the Driving Behavior of a

Vehicle”, by Peter Wanke, Torsten Herrman, Mario Roszyk, Thorsten Wickenhofer, Matthias
Pank, and Jochen Konanz, assigned to Continental Teves AG & Co oHG, March 23,
29United Nations Economic Commission for Europe (UN/ECE) Regulation No. 13-H (2017) with

Annexes 1 to 9, “Technical Requirements, Test Methods, and Limit Values for the Braking
Systems of Passenger Cars and Light Commercial Vehicles”, par 2.17.2.
30 United Nations Economic Commission for Europe of the (UN/ECE) Regulation No 140

“Uniform provisions concerning the approval of passenger cars with regard to Electronic
Stability Control (ESC) Systems”, par 6.2.2, 6.2.4.
31 Ibid, Note 22, 2018/1592, par 6.2.3.
32 Ibid, Note 22, 2018/1592, par 9.10.1.
33 United Nations Economic Commission for Europe (UN/ECE) Regulation No. 13-H (2017)

with Annexes 1 to 9, “Technical Requirements, Test Methods, and Limit Values for the Braking
Systems of Passenger Cars and Light Commercial Vehicles”, par 5.2.22.4.
34 USA regulation 49 CFR § 571.126 Federal Standard No. 126; “Electronic stability control

systems for light vehicles”, par S5.1.3.

35 “ID4EV -- Intelligent Dynamics for Electric Vehicles”, a study on the requirements and

specifications for electric vehicles derived from user needs and SOTA analysis performed under
European research project co-funded by the European Commission under the 7th Framework
programme, p 33.
36 Ibid, Note 27, p77.
37 Ibid, Note 27, p77, par j.
38 Ibid, Note 27, p77, par j.
39 “Investigation Into Regenerative Braking Systems” by B. J. Robinson, C. Visvikis, T. Gibson,

and I. Knight of the Transport Research Laboratory, published project report No. PPR582 for
the Department for Transport, Transport Technology and Standards, United Kingdom, 2011,
p10.
40 Ibid, Note 31, p12.
41 https://forums.tesla.com/de_CH/node/87902. This is the German Tesla forum.
42 https://teslamotorsclub.com/tmc/threads/another-sudden-acceleration.157549/
43 https://teslamotorsclub.com/tmc/threads/regenerative-braking-safety-issue.52053/page-2
44 https://fordfusionhybridforum.com/topic/9866-regen-braking-issue-when-turning/page/3/
45 https://teslamotorsclub.com/tmc/threads/reduced-regen-with-lateral-gs.115479/

Tesla Regen, Brakes and 65 R. Belt

Sudden Acceleration 1 June 2020
46https://www.courtlistener.com/recap/gov.uscourts.cacd.770392/gov.uscourts.cacd.770392.1.

0.pdf
47 https://insideevs.com/news/364540/video-bmw-i3-crash-building/
48 https://www.facebook.com/permalink.php?story_fbid=1928733057185174&id=2392958727

95576
49 https://www.bmwblog.com/2014/06/23/living-bmw-i3-2000-miles-dislikes/
50 https://insideevs.com/reviews/336237/bmw-i3-sport-one-month-review/
51 https://www.autosafety.org/dr-ronald-a-belts-sudden-acceleration-papers/
52 https://autotechreview.com/cover-stories/electromechanical-brake-booster-for-all-drive-

concepts-automated-driving-2
53 https://www.greencarcongress.com/2018/05/20180522-cr.html
54 https://www.forbes.com/sites/samabuelsamid/2018/05/30/rapid-tesla-software-updates-

are-great-but-hide-more-fundamental-problems/#2d5802e657de
55 http://www.news.com.au/national/hyundai-models-in-voluntary-recall-over-faulty-brake-

light-switch/story-fncynjr2-1226591850472
56 https://www.edmunds.com/car-news/2009-11-hyundai-accent-recalled-for-brake-switch-

problem.html
57 https://www.goauto.com.au/news/kia/kia-acts-on-faulty-brake-switch/2013-09-

26/21715.html
58 https://www.foxbusiness.com/features/subaru-announces-it-biggest-recall-ever-over-brake-

lights
59 https://www.goauto.com.au/news/kia/kia-acts-on-faulty-brake-switch/2013-09-

26/21715.html
60 NHTSA Campaign Number: 19V418000, https://www.citybike.com/news-clues-

rumors/suzuki-gsx250r-recall-brake-light-failure/
61http://pressroom.toyota.com/releases/customer+faqs+toyota+venza+camry+stop+lamp+swit

ch+recall.htm
62 https://www.consumeraffairs.com/news/gm-recalls-nearly-25-million-vehicles-with-

electrical-problems-051614.html
63 http://www.huffingtonpost.com/2014/05/29/gm-recall-delay-documents_n_5412979.html
64 http://z71tahoe-suburban.com/iboard/lofiversion/index.php?t38495.html
65 https://www.equipmentworld.com/ford-issues-recalls-for-f-150-due-to-brake-light-and-seat-

issues/
66 http://www.f150forum.com/f10/brake-light-delay-153363/
67 http://www.bimmerforums.com/forum/showthread.php?1154981-Brake-Light-Delay
68 https://teslamotorsclub.com/tmc/threads/brake-lights-not-coming-on-w-

proof.122555/#post-2872074
69
https://www.courtlistener.com/recap/gov.uscourts.cacd.770392/gov.uscourts.cacd.770392.1.
0.pdf

Tesla Regen, Brakes and 66 R. Belt

Sudden Acceleration 1 June 2020