» BOOKSHELF

arrangements and applications. The remainder of the text

I EEE Control Systems welcomes suggestions for books to

is organized into four parts: geometry of robots, dynamics
be reviewed in this column. Please contact either Scott R.
and motion planning, control of manipulators, and control
Ploen, Hong Yue, or Thomas Schön, associate editors for book
of underactuated systems.
reviews.
Part 1, consisting of Chapters 2–4, deals with the geom-
etry of rigid motions and kinematics of open-chain manip-
ulators. Chapter 2 presents the mathematical formalism
ROBOT MODELING describing rigid motions, that is, rotations, translations,
AND CONTROL, and homogeneous transformations. Although this chapter
SECOND EDITION briefly introduces the notion of exponential coordinates
and twists for general rigid motions, it emphasizes the use
by MARK W. SPONG,
SETH HUTCHINSON,
of homogeneous transformations in the subsequent devel-
and M. VIDYASAGAR opment of robot kinematics.
Chapter 3 investigates the forward kinematics problem
Reviewed by Alireza Mohammadi and provides a systematic way for solving these problems
using the Denavit–Hartenberg (DH) representation which,

T
he first edition of this book unlike the product of exponentials formula [1], relies on
inspired many research- assigning local link frames. Nevertheless, emphasizing the
ers and educators working DH representation keeps the required mathematical level
at the intersections of robotics, accessible to an undergraduate audience without sacrific-
Wiley, 2020, control theory, and computer ing the depth and rigor of the material in later chapters.
ISBN: 978-1-119-52399-4, vision from the early days of its Chapter 4 discusses velocity kinematics and the manip-
608 Pages, US$145. publication. This second edi- ulator Jacobian. Both the geometric Jacobian in cross-prod-
tion, published 14 years after its uct form and the analytical Jacobian for task space control
predecessor, maintains the central perspective that robots are are introduced. The order of presentation is different from
machines that transform sensing into action through feed- the first edition as it gives the velocity kinematics first to
back control, with the goal of manipulation of objects or loco- support the inclusion of a new section in Chapter 5 by
moting in their ambient environment. Given its central focus focusing on numerical inverse kinematics algorithms that
on feedback control, along with its comprehensive treatment utilize the Jacobian.
of linear, nonlinear, and geometric control algorithms for ro- Chapter 5 starts with solving the inverse kinematics problem
botic manipulators and mobile robots, this book holds a spe- using the geometric approach, which is particularly suitable for
cial place among other robotics textbooks [1]–[5]. manipulators with spherical wrists. Both closed-form solutions
In addition to updating the material on motion plan- for the inverse kinematics of some of the most common manip-
ning, vision, and vision-based control to reflect state-of- ulator designs and solutions based on numerical search algo-
the-art changes in the field, the current edition contains rithms are presented. Specifically, numerical techniques based
two new chapters dedicated to the emerging topic of under- on the Jacobian inverse method (using a Newton–Raphson
actuated robotic mechanisms, including a significant search) and the Jacobian transpose procedures (using gradient
extension of the coverage of mobile robots. This edition descent approaches) are emphasized. Due to these newly
also contains two new appendices dedicated to the topics of added optimization-based algorithms, a new appendix on opti-
optimization and camera calibration. mization has been added to the second edition.
Part 2, consisting of Chapters 6 and 7, covers dynamics
CONTENTS and motion planning. Chapter 6 provides a detailed
Chapter 1 establishes basic terminology, presents a brief account of robot dynamics using both the Euler–Lagrange
history of robotics, and discusses common robot kinematic equations and the recursive Newton–Euler formulation.
The Lagrangian structure of the resulting equations of
Digital Object Identifier 10.1109/MCS.2021.3122271 motion, which are extensively used for the design and anal-
Date of current version: 18 January 2022 ysis of control algorithms in Part 3, are discussed in detail.

126 IEEE CONTROL SYSTEMS » FEBRUARY 2022 1066-033X/22©2022IEEE

The coverage of this material, although concise, provides in the workspace for controlling robots. Throughout the
the necessary dynamics preliminaries for electrical and treatment of this material, this book provides derivations
computer engineering students who might have only a based on rigorous stability proofs. Aspects of vision such
minimal exposure to this topic. as imaging geometry and feature extraction, which are
Chapter 7 introduces the topics of path and trajectory plan- beneficial for vision-based robot control, are also treated in
ning, along with several of the most widely used techniques detail. After developing the differential kinematics formal-
for motion planning and obstacle avoidance (including the ism and discussing how it relates camera motion to changes
method of artificial potential fields, randomized algorithms, in extracted features, the main concepts of visual servo
and probabilistic road map approaches). For planar path- control are also examined. A new appendix on camera cali-
planning problems, this chapter also presents the concepts of bration has also been added to this edition.
visibility graphs, generalized Voronoi diagrams, and trape- Chapter 12 provides a tutorial overview of geometric non-
zoidal decompositions. The polynomial spline-interpolation linear control and the method of feedback linearization for
formalism is used to present the trajectory-generation prob- nonlinear systems. After a detailed derivation of the necessary
lem in joint space. This formalism results in trajectory genera- and sufficient conditions for local feedback linearization of
tion using cubic and quintic polynomials as well as the single-input, single-output nonlinear systems, this methodol-
derivation of trapezoidal-velocity trajectories for interpolation ogy is applied to the flexible joint control problem. Addition-
in joint space. ally, the chapter provides a brief overview of dynamic output
Part 3, consisting of Chapters 8–12, deals with the con- feedback control using nonlinear observers with output injec-
trol of manipulators. This part, as well as Part 4, distin- tion, which is a generalization of the topic of linear observers
guish the book from its other robotic textbook counterparts presented in Chapter 8. The flexible joint control problem pre-
[1]–[5] because of its extensive treatment of control algo- sented in this chapter serves as an entry point for delving into
rithms for robotic manipulators and mobile robots. the topic of underactuated robotics in the final part of this book.
Chapter 8 introduces the topic of independent joint con- Part 4, consisting of Chapters 13 and 14, is a completely
trol for robots where linear models and linear control tech- new addition to the second edition and treats control prob-
niques such as proportional-integral-derivative control and lems for underactuated robots and nonholonomic systems.
state-space-based algorithms play a crucial role. Accord- Chapter 13 is motivated by recent advances in robotic loco-
ingly, these linear control methods are used to address set- motion. With applications to bipedal and gymnastic robots
point regulation, trajectory tracking, and disturbance in mind, this chapter presents the central ideas of feedback
rejection in robot motion control problems. Following the transformation to normal forms and partial feedback lin-
presentation of linear styles for robot control, this chapter earization for controller design. The chapter also contains
introduces the concept of feedforward control, with a spe- discussions on energy-based and passivity-based control
cific emphasis on the method of computed torque control methods for underactuated robotic mechanisms.
used for nonlinear disturbance rejection and tracking time- Chapter 14 provides a significant extension to the topic
varying reference trajectories. of wheeled mobile robots, which was only briefly reviewed
Chapter 9 discusses nonlinear and multivariable con- in the first edition. These classes of nonholonomic systems
trol, including an extensive summary of the breakthroughs do not lend themselves to the control design methodologies
in robot control during the late 1980s and early 1990s. presented in previous chapters. Accordingly, two funda-
Although broadly covering the most common robust and mental results from geometric control theory, Chow’s theo-
adaptive control techniques for robot manipulators, this rem and Brockett’s theorem, are presented. The former
chapter maintains its faithfulness to simple derivations. A theorem provides conditions for controllability, while the
whole section is dedicated to the topic of passivity-based latter provides the necessary conditions for stabilizability
control and can serve as motivation for further studies in of mobile robots. Furthermore, suitable control and motion
port-Hamiltonian approaches to robot control [6]. Corre- planning algorithms such as dynamic input extension,
sponding to the emergence of optimization-based control- steering using sinusoids, and differential flatness-based
lers for robotic control systems, the chapter concludes with control for mobile robots are thoroughly discussed.
a newly added section on torque optimization using point-
wise-quadratic, programming-based solutions. Appendix B SUMMARY
on Lyapunov stability theory makes the treatment of non- Since the first edition of the book, several textbooks and
linear control largely self-contained. research monographs such as those in [1]–[5] focusing on
Chapter 10 treats the force control problem using imped- robotics have been published. Nevertheless, this updated
ance and hybrid control methods. To allow simultaneous second edition has maintained its central place in the field
control of robot force and motion, this chapter presents the of robotic research and education due to its authoritative
lesser-known hybrid impedance control scheme. treatment of robotic feedback control. In particular, Parts 3
Chapter 11 is dedicated to visual servo control, namely, and 4 provide an unparalleled comprehensive treatment
employing feedback from cameras mounted on the robot or of linear, nonlinear, geometric, and visual servo control

FEBRUARY 2022 « IEEE CONTROL SYSTEMS 127

methods for robotic manipulators, and mobile and under- 2016 to 2018, he was a postdoctoral research associate with
actuated robots. the Locomotor Control Systems Laboratory, the University
Finally, when teaching a course from this book, other of Texas at Dallas, Richardson, Texas, USA. He is currently
robotics textbooks can be consulted for complimentary an assistant professor with the Department of Electrical
treatments of various topics. The material on forward and and Computer Engineering, the University of Michigan–
inverse kinematics of closed chains as well as grasping and Dearborn, Dearborn, Michigan, 48128, USA. His current re-
manipulation using form/force closures from [1] compli- search interests include collaborative robotics, cybersecu-
ment Chapters 3–5 and 10. The complete set of Matlab rity of in-vehicle networks, mechatronics, nonlinear control
codes provided in [5] can be used to further enrich a robot- theory, and hybrid dynamical systems.
ics course. In teaching undergraduate robotics courses, [2]
provides a nice complement to Chapter 6 on the Newton– REFERENCES
Euler formulation of robot dynamics. In courses where [1] K. M. Lynch and F. C. Park, Modern Robotics. New York, NY, USA: Cam-
bridge Univ. Press, 2017.
path planning is investigated in more depth, [7] provides a [2] J. J. Craig, Introduction to Robotics: Mechanics and Control, 4th ed. Upper
complimentary reference to the material from Chapter 7. Saddle River, NJ, USA: Pearson Education, 2018.
[3] R. Siegwart, I. R. Nourbakhsh, and D. Scaramuzza, Introduction to Au-
tonomous Mobile Robots. Cambridge, MA, USA: MIT Press, 2011.
REVIEWER INFORMATION [4] R. M. Murray, Z. Li, and S. S. Sastry, A Mathematical Introduction to Robotic
Alireza Mohammadi (amohmmad@[Link]) received Manipulation. Boca Raton, FL, USA: CRC Press, 1994.
his M.S. degree from the University of Alberta, Edmonton, [5] P. Corke, Robotics, Vision and Control: Fundamental Algorithms in MAT-
LAB®, 2nd ed. Cham: Springer Nature Switzerland AG, 2017.
Alberta, Canada, in 2011 and the Ph.D. degree in electri- [6] C. Secchi, S. Stramigioli, and C. Fantuzzi, Control of Interactive Robotic
cal and computer engineering from the University of To- Interfaces: A Port-Hamiltonian Approach. Cham: Springer Nature Switzerland
ronto, Toronto, Ontario, Canada, in 2016. During his Ph.D. AG, 2007.
[7] H. M. Choset et al., Principles of Robot Motion: Theory, Algorithms, and
studies, he collaborated with the Norwegian Centre for Implementation. Cambridge, MA, USA: MIT Press, 2005.
Autonomous Marine Operations and Systems on locomo-
tion control of ground and swimming snake robots. From 

» PEOPLE IN CONTROL (continued from p. 53)

opportunities exist in the control field

Profile of Jie Chen in modern society. The control of com-
• Current position: president, Tongji University, China; director, State Key Labo- plex systems from social networks to
ratory of Intelligent Control and Decision of Complex Systems, China; vice artificial systems like autonomous ro-
president, Chinese Association of Automation (2013–present); vice president, bots and their interconnection networks
Chinese Association for Artificial Intelligence (2019–present). still has a long way to go.
• Contact information: Tongji University, 1239 Siping Road, Shanghai, 200092,
China, chenjie206@[Link], chenjie@[Link]. Q. What are some of your interests
• Notable awards: Distinguished Young Scholar, Natural Science Foundation of and activities outside of your profes-
China (2009); Changjiang Scholar Distinguished Professor, Chinese Ministry of sional career?
Education (2011); Ho Leung Ho Lee Foundation Prize (2018); National Natural Jie: I love reading and riding. A
Science Award of China (2014); National Science and Technology Progress Chinese proverb says, “A man needs
Award of China (2009 and 2011). to read ten thousand books and travel
ten thousand miles.” Different kinds
of books open my mind. Riding keeps
Q. What are some of the most prom- a pursuit in our society and daily lives me fit. Both drive me to move forward.
ising opportunities you see in the for a long time, and it will be in the fu-
control field? ture. We have been creating various Q. Thank you for your comments.
Jie: As control exists almost every- machines or devices to benefit humans Jie: Thank you very much for the
where, it has close connections to di- and facilitate social ­d evelopment. opportunity.
verse disciplines. Automation has been I have no doubt that many promising 

128 IEEE CONTROL SYSTEMS » FEBRUARY 2022