Sequencing (Models)

Job Sequencing What is best sequence : Based on objective Assumptions : (1) One operation on one machine at a time.

(2) Processing times known and constant.

(3) Processing times not depending on sequence.

(4) Idleness of job between any two machines = 0.

(5) Operation started must be completed. (6) Only one machine of each type is available.

Categories :
[1] n jobs 1 machine

FSP (Flow Shop Problem)

[2] n jobs 2 machines JSP (Job Shop Problem)

[3] n jobs 3 machines

[4] n jobs m machines

[5] 2 jobs m machines

[1] n job 1 machine

Data for this case :
Jobs (Ji) 1 2 3 4 5 6 7 8

Processing Time (Pi)

Due days (Di) Priority (wi)

5
15 1

8
10 2

6
15 4

3
25 5

10
20 3

14
40 8

7
45 6

3
50 7

If sequence is 1-2-3-4-5-6-7-8 then

Ji Pi Fi 1 5 5 2 8 13 3 6 19 4 3 22 5 10 32 6 14 46 7 7 53 8 3 56

Mean Flow Time (MFT) = Fi / No. of jobs = 246/8 = 30.75

Hence, one of the criterion of Optimization is MFT.

If sequence is 1-2-3-4-5-6-7-8 then

Ji
Pi Fi Di Li

1
5

2
8

3
6 19 15 4

4
3 22 25 -3

5
10 32 20 12

6
14 46 40 6

7
7 53 45 8

8
3 56 50 6

5
15 -10

13
10 3

Lmax = 12 for 5th job,

No. of late jobs = 6

Lavr = Li / No. of jobs = 26/8 = 3.25

Hence, the other criteria of Optimization are Maximum Lateness of a Job as well as Number of Late Jobs.

Jobs

Processing Time (Pi)

Due days (Di) Priority (wi) Pi * Wi

5
15 1 5

8
10 2 16

6
15 4 24

3
25 5 15

10
20 3 30

14
40 8 112

7
45 6 42

3
50 7 21

Pi * Wi is weighted processing time

Hence, the other criterion of Optimization is Weighted Processing Time.

Criteria of Optimization (A) Mean Flow Time (MFT)

(B) Maximum lateness

(C) Number of late jobs (D) Priority

If sequence is 1-2-3-4-5-6-7-8 then

Ji
Pi Fi Di Li

1
5

2
8

3
6 19 15 4

4
3 22 25 -3

5
10 32 20 12

6
14 46 40 6

7
7 53 45 8

8
3 56 50 6

5
15 -10

13
10 3

MFT = Fi / No. of jobs = 246/8 = 30.75

Lavr = Li / No. of jobs = 26/8 = 3.25 Lmax = 12 for 5th job, No. of late jobs = 6

(A) To minimize MFT, Shortest Processing Time (SPT) rule is applied.

Ji Pi 1 5 2 8 3 6 4 3 5 10 6 14 7 7 8 3

Applying SPT rule, the optimal sequence to minimize MFT will be : 4-8-1-3-7-2-5-6

To get MFT, Lavr, Lmax and No. of late jobs :

Ji Pi Fi Di

4 3 3 15

8 3 6 10

1 5 11 15

3 6 17 25

2 7 24 20

7 8 32 40

5 10 42 45

6 14 56 50

Li

-12

-4

-4

-8

-8

-3

MFT = Fi / No. of jobs = 191/8 = 23.87

Lavr = Li / No. of jobs = 29/8 = 3.62

Lmax = 6 for 6th job, No. of late jobs = 2

(B) To minimize maximum lateness, Shortest Due Days (SDD) rule is applied.
Ji Pi Di 1 5 15 2 8 10 3 6 15 4 3 25 5 10 20 6 14 40 7 7 45 8 3 50

Applying SDD rule, the optimal sequence to minimize maximum lateness will be : 2-1-3-5-4-6-7-8 To get MFT, Lavr, Lmax and No. of late jobs :

To minimize maximum lateness, Shortest Due Days (SDD) rule is applied.

Ji Pi 2 8 1 5 3 6 5 10 4 3 6 14 7 7 8 3

Fi Di
Li

8 10
2

13 15
2

19 15
4

29 20
9

32 25
7

46 40
6

53 45
8

56 50
6

MFT = Fi / No. of jobs = 256/8 = 32 Lavr = Li / No. of jobs = 36/8 = 4.50 Lmax = 9 for 5th job, No. of late jobs = 6

(C) To minimize number of late jobs : Sequence for minimum (maximum lateness)
Ji Pi Fi 2 8 8 10 2 1 5 13 15 2 3 6 19 15 4 5 10 29 20 9 4 3 32 25 7 6 14 46 40 6 7 7 53 45 8 8 3 56 50 6

Di
Li

Lmax = 9 for 5th job, No. of late jobs = 6 The first job with lateness is 3rd Job (Job No. 3) Out of the first three jobs Pij is Max. (=8)for 1st job (Job No. 2) For Minimizing late jobs remove Job2 (Pi = 8)

Ji Pi Fi Di

1 5 5 15 10

3 6 11 15 4

5 10 21 20 1

4 3 24 25 1

6 14 38 40 2

7 7 45 45 0

8 3 48 50 2

Li

Lmax = 1 for 5th job,

No. of late jobs = 1

The first job with lateness is 3rd Job (Job No. 5) Out of the first three jobs Pij is Max. (=10)for 3rd job (Job No. 5) For Minimizing late jobs remove Job5 (Pi = 10)

To minimize number of late jobs :

Sequence for minimum (maximum lateness)

Ji Pi Fi Di Li 1 5 5 15 10 3 6 11 15 4 3 14 25 6 14 28 40 7 7 35 45 8 3 38 50 12

4 11 12 10

No job is late. Hence, the optimal sequence to minimize late jobs will be 1-3-4-6-7-8-2-5 or 1-3-4-6-7-8-5-2

To minimize number of late jobs :

Sequence for minimizing No. of late jobs

Ji Pi Fi Di Li 1 5 5 15 10 3 6 11 15 4 3 14 25 6 14 28 40 7 7 35 45 8 3 38 50 12 2 8 46 10 36 5 10 56 20 36

4 11 12 10

Lmax = 36 for 2nd & 5th job,

No. of late jobs = 2

Lavg = Li / No. of jobs = 13/8 = 1.62

(D) Based on priority

Jobs 1 2 3 4 5 6 7 8

Processing Time (Pi)

Due days (Di) Priority (wi) Pi * Wi

5
15 1 5

8
10 2 16

6
15 4 24

3
25 5 15

10
20 3 30

14
40 8 112

7
45 6 42

3
50 7 21

Pi * Wi is weighted processing time Based on weighted processing time the optimal sequence will be :

14283576

[1] Jobs Pi di 1 5 4 2 8 25

Exercise
3 3 6 4 10 36 5 4 10 6 7 28

wi
[2] Jobs 1 2

4
[3]

Pi 10 7

di 11 9

wi 2 4

Jobs 1 2

Pi 3 6

di 6 20

wi 4 1

3 4 5 6

3 15 6 9

35 29 42 50

1 3 6 5

3 4 5

9 5 7

30 12 25

5 3 2

[2] n jobs 2 machines

Flow Shop Problem (FSP)

Job Shop Problem (JSP)

(a) Flow Shop Problem (FSP)

Jobs 1 M1 8 M2 2

2
3 4 5 6

5
3 11 1 12

10
9 6 7 4

Elapsed Time for a Sequence is the duration from starting of 1st Job to the end of last Job.
The criterion of Optimization is Elapsed Time.

To get Sequence to Minimize Elapsed Time, Johnsons Rule is applied.

Jobs 1 2 3 M1 8 5 3 M2 2 10 9

4
5

11
1

6
7

6
Applying Johnsons rule :

12
5

Flow Shop Problem (FSP) Jobs 1 2 3 4 5 6 M1 8 5 3 11 1 12 5 M2 2 10 9 6 7 4 1

Flow Shop Problem (FSP) Jobs 1 2 3 4 5 6 M1 8 5 3 11 1 12 5 3 M2 2 10 9 6 7 4 1

Flow Shop Problem (FSP) Jobs 1 2 3 4 5 6 M1 8 5 3 11 1 12 5 3 M2 2 10 9 6 7 4 6 1

Flow Shop Problem (FSP) Jobs 1 2 3 4 5 6 M1 8 M2 2 10 9 6 7 4 3 2 6 1

5
3 11 1 12 5

Flow Shop Problem (FSP) Jobs 1 2 3 4 5 6 M1 8 5 3 11 1 12 5 3 2 M2 2 10 9 6 7 4 4 6 1

Hence, Optimal Job Sequence is : 5 3 2 4 6 1

How to get Elapsed Time (T) for any sequence ?

Elapsed Time for a Sequence is the total time from starting of first job to the completion of last job in a sequence.
To get Elapsed (Makespan) Time for sequence 1-2-3-4-5-6
1 M1 8 M2 10 0

Jobs
1 2 3 4 5 6

M1
8 5 3 11 1 12

M2
2 10 9 6 7 4

Jobs

M1

M2

1
2
1 2

8
5 3 11 1 12

2
10 9 6 7 4

3 4
13

M1
8

5 6

10

23

Jobs 1 2
1 2 3

M1 8 5 3 11 1 12

M2 2 10 9 6 7 4

3 4

M1
8 13 16

5 6

10

23

32

Jobs 1 2
1 2 3 4

M1 8 5 3 11 1 12

M2 2 10 9 6 7 4

3 4
27

M1
8 13 16

5 6

10

23

32

38

Jobs 1 2
1 2 3 4 5

M1 8 5 3 11 1 12

M2 2 10 9 6 7 4

3 4

M1
8 13 16 27 28

5 6

10

23

32

38

45

Jobs 1 2 1 2 3 4 5 6 3 4

M1 8 5 3 11

M2 2 10 9 6

M1
8 13 16 27 28 40

5
6

1
12

7
4

10 T = 49

23

32

38

45 49

Elapsed Time = 49 Idleness of M1 = 9

Idleness of M2 = 11

For Optimal Job Sequence : 5 3 2 4 6 1 Find elapsed time for this optimal sequence. Ans. : T = 42 days Idleness of M1 = 2 days, Idleness of M2 = 4 days

Exercise
Find the optimal sequence of jobs and elapsed time for optimal sequence for the following n jobs 2 machines (FSP) problems. All jobs pass through M1 M2 sequence. Timings are given in days.
Jobs 1 2 3 4 5 6

M1
M2

7
2

5
4

1
3

10
7

6
11

3
8

(b) Job Shop Problem (JSP)

Job Machine for Processing time for O1 O2 O1 O2
1 2 3 4 5 6 M1 M1 M2 M1 M2 M2 M2 -M1 M2 -M1 2 4 3 6 3 7 8 6 3 4 [ M1 ] [ 2 ] [ M2 ] [ 5, 8 ] [ M1 M2 ] [ 1, 4, 7, 9 ]

[ M2 M1 ] [ 3, 6, 10 ]
Jacksons Rule : On M1 [ M1 M2 ] [ M1 ] [ M2 M1 ] ..

7
8 9

M1
M2 M1

M2
-M2

9
2 4

- On M2 [ M2 M1 ] [ M2 ] [ M1 M2 ] .. 7

10

M2

M1

To get Optimal Sequence of [ M1 M2 ] jobs :

[ M1 M2 ] [ 1, 4, 7, 9 ]
Job Machine for Processing time for O1 O2 O1 O2 1 4 7 9 M1 M1 M1 M1 M2 M2 M2 M2 2 6 9 4 8 3 5 7

Applying Johnsons rule :

[ M1 M2 ] [ 1, 4, 7, 9 ] 1, 9, 7, 4

To get Optimal Sequence of [ M2 M1 ] jobs :

[ M2 M1 ] [ 3, 6, 10 ]
Job Machine for Processing time for O1 O2 O1 O2 3 M2 M1 3 6

6
10

M2
M2

M1
M1

7
5

4
8

Applying Johnsons rule :

10

[ M2 M1 ] [ 3, 6, 10 ] 3, 10, 6

[ M1 ] [ 2 ] 2 [ M2 ] [ 5, 8 ] 5, 8 [ M1 M2 ] [ 1, 4, 7, 9 ] 1, 9, 7, 4 [ M2 M1 ] [ 3, 6, 10 ] 3, 10, 6 Jacksons Rule : On M1 [ M1 M2 ] [ M1 ] [ M2 M1 ] .. On M2 [ M2 M1 ] [ M2 ] [ M1 M2 ] .. Hence Sequence on M1 : 1-9-7-4-2-3-10-6 Sequence on M2 : 3-10-6-5-8-1-9-7-4

Exercise
Find the optimal solution for the following Job-Shop problem. Timings are given in days.

Jobs 1 2 3 4 5 6 7 8 9

Machine for Operation1 Operation2 M1 M1 M1 M1 M2 M2 M2 M2 M2 M2 M2 M1 M1 M1

Operation1 4 3 4 5 1 1 7 3 6

Operation2 6 2 2 8 7

10

M1

M2

[3] n jobs 3 machines : Flow Shop Problem

Jobs 1 2 3 4 5 A 5 7 9 6 8 B 4 3 1 2 5 C 3 4 2 3 3

Min = 5 Max = 5 Min = 2

At least any one condition out of the following has to be satisfied to convert this problem into 2 m/cs problem.
(1) Maximum of middle (2) Maximum of middle Minimum of first column Minimum of last column

Jobs 1 2 3 4 5

A 5 7 9 6 8

B 4 3 1 2 5

C 3 4 2 3 3

Jobs
1 2 3 4 5

G
9 10 10 8 13

H
7 7 3 5 8

Jobs 1 2 3 4 5 Applying Johnsons rule :

G 9 10 10 8 13

H 7 7 3 5 8

2
OR

Hence, optimal solution of original problem is 5-1-2-4-3 OR 5-2-1-4-3

[4] n jobs m machines : Flow Shop Problem

Jobs M1 M2 M3 M4 M5

1 2 3 4

9 8 7 10

7 8 6 5

4 6 7 5

5 7 8 4

11 12 10 8

Min = 7 Max = 8 Min = 8 At least any one condition out of the following has to be satisfied to convert this problem into 2 m/cs problem. (1) Maximum of middle Minimum of first column (2) Maximum of middle Minimum of last column

Jobs 1 2 3 4 Jobs

M1 M2 M3 M4 M5 9 8 7 10 7 8 6 5 G 4 6 7 5 5 7 8 4 H 11 12 10 8

1 2 3 4

25 29 28 24

27 33 31 22

Jobs 1 2 3 4

G 25 29 28 24

H 27 33 31 22

Applying Johnsons rule :

Hence, optimal solution of original problem is 1-3-2-4

Exercise
Get the optimal solution for following n-jobs m-machines problem. Timings are given in days.
Jobs 1 2 3 4 M1 18 17 11 20 M2 8 6 5 4 M3 7 9 8 3 M4 2 6 5 4 M5 10 8 7 8 M6 25 19 15 12

[5] 2 jobs m machines Problems :

Job1 C(3), A(3), D(4), B(2), E(6) Job2 B(2), A(4), C(4), D(2), E(4)

[5] 2 jobs m machines Problems :

Job1 C(3), A(3), D(4), B(2), E(6) Job2 B(2), A(4), C(4), D(2), E(4)

16 E

J2 E

12 D

8 C

C T = 18+16 = 34

4 A
B C

4 A

8 D

B 12 B E

16

J1

16 E

J2 E

12 D

8 C

C T = 16+18 = 34

4 A
B C

4 A

8 D

B 12 B E

16

J1

16 E

12,16 E

L18,16

12 D

12,12

D
7,10 10,10

8 C

C
3,6

T = 18+7 = 25

4 A
B

3,3

A 6,3
B 12 B E

4 C A

8 D

16

16 E

12,16
E

18,16

12 D

12,12
10,10

8 C

C
6,6

T = 18+8 = 26

4 A
B

A
3,2 6,2 4 C A 8 D B 12 B E 16

16 E 10,12

12,16
E

18,16

12 D

12,12
10,10

8 C

C
6,6

T = 18+9 = 27

4 A
B

A
3,2 6,2 4 C A 8 D B 12 B E 16

16 E E 12 D 16,12

18,12

8 C

T = 18+4 = 22

4 A
B

A
3,2 6,2 4 C A 8 D B 12 B E 16

Hence, minimum time in which both the jobs

can be completed = 22 days.

Scheduling is as per the blue path shown w.r.t. 22 days.

Exercise
Solve the following 2-jobs, m-machines problems. Find graphically minimum time required to complete both the jobs. Timings are given in days. [1] Job 1 Job 2 [2] Job 1 Job 2 Seq. of m/c Time Seq. of m/c Time A 2 C 4 B 3 A 5 C 4 D 3 D 6 E 2 E 2 B 6 Seq. of m/c Time A 3 B 4 C 2 D 6 E 2

Seq. of m/c Time

B 5

C 4

A 3

D 2

E 6

Thank you
For any Query or suggestion :
Contact : Dr. D. B. Naik Professor & Head, Training & Placement, S. V. National Institute of Technology, Ichchhanath, Surat 395 007 (Gujarat) Email : [email protected] Ph. : 0261-2201540, 2255225 (O)