You are on page 1of 31

PSG COLLEGE OF TECHNOLOGY

COIMBATORE – 641004

FACILITY LAYOUT OPTIMIZATION OF A MANUFACTURING


INDUSTRY USING GENETIC ALGORITHM AND SIMULATED
ANNEALING

Guided by Presented by
Dr A PRABUKARTHI ARAVIND NARAYANAN V
ASSISTANT PROFESSOR 19MF35
Dept. of Mechanical Engineering ME INDUSTRIAL ENGINEERING

05/17/2021 1
GUIDE APPROVAL

SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL ENGG. 2
COMPANY PROFILE

• Name of the enterprise : DEV MAX ENERGY

• Services offered : Boiler, pipeline fabrication and erection, heat pump supply and
installation and service provider of MS tanks, heat exchangers

• Location : Chennai

SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL ENGG. 3
PROBLEM STATEMENT
• The industry adapts a type of layout in which different machineries of different products are
collectively arranged in a single factory floor and their positions are fixed.

• Since this is an ETO type, only one product is being manufactured at a time. Because of this type
of layout, Workers and material handling equipment are forced to move around some machines
which are not used in the process, to get to the desired place. This causes unnecessary
movements which increases
Material handling cost
Production time

SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL
18MN72 ENGG. 4
NEED OF THE WORK
• Layout optimization is performed to understand the industry’s utilization rate and material handling
efficiency.

• This study focuses purely on material handling aspect, when implemented will provide an better
outcome in terms of movement of materials and the cost associated with it.

• The methodology involving meta heuristic procedures, will give an extended insight on how to solve
a facility layout problem mathematically providing quantifiable results.

SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL
18MN72 ENGG. 5
LITERATURE SUMMARY
• From the literature study, it has been proven that heuristic algorithms can be used to solve any
type of facility layout problems.

• Formulation of equations with respect to genetic algorithm and simulated annealing can be derived
using a simple product facility layout and the same can be used to obtain a optimal layout.

• Multi objective layout problems can be solved using a single heuristic equation which yields optimal
results.

• The large-scale facility layout optimization algorithm and the combination scheme with fuzzy logic
and AHP can be chosen as the direction of further research.
SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL
18MN72 ENGG. 6
OBJECTIVES
1.To provide optimal cost of the layout using genetic algorithm and simulated annealing and map the
final layout in order to restrict unnecessary movements.

2.To compute the process time using discrete event simulation for the product.

3.To reduce the overall production time by comparing the existing facility layout with the optimal layout
obtained by using a operation process chart.

SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL
18MN72 ENGG. 7
METHODOLOGY

SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL
18MN72 ENGG. 8
DESCRIPTION OF THE MAIN WORK

SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL
18MN72 ENGG. 9
FLOW PROCESS CHART (MS TANK)

SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL
18MN72 ENGG. 10
SIMULATION USING ARENA

Process time(Using arena) = 111 mins


Total Material handling time = 42 mins
Overall
production time = 153 mins

SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL
18MN72 ENGG. 11
FORMULATION
• CRAFT algorithm is used to formulate the objective function

Area legend Department legend


1 – (8 X 6) feet 1 – Rolling and bending machine
2 – (8 X 6) feet 2 – Cutting machine
3 – (8 X 4) feet 3 – Drilling machine
4 – (8 X 6) feet 4 – Welding machine
5 – (10 X 8) feet 5 – Assembly workstation 1
6 – (10 X 8) feet 6 – Assembly workstation 2

SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL
18MN72 ENGG. 12
FORMULATION
Coordinates are:
• Next step is to find the distance matrix ( Dij ) for (x1,y1) = (4, 11)
(x2,y2) = (12,11)
which,
(x3,y3) = (22,11)
Dij =
(x4,y4) = (28,11)
(x5,y5) = (5,4)
(x6,y6) = (15,4)

SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL
18MN72 ENGG. 13
FORMULATION
 • Next step is to find the initial total cost of the layout using the formula,

Tij = (1)

Tij = Rs. 14856

• Assume every value in cost matrix (Cij ) as “x”. Then the initial total cost using equation (1) will be

Tij for finding the optimal cost

• Assume every value in distance matrix (Dij ) as “y”. Then the initial total cost using equation (1) will
be

Tij for finding the optimal distance


SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL
18MN72 ENGG. 14
FORMULATION
 • Overall cost of the layout is

which is the objective function

• Objective function is to minimize the overall layout cost

Min

Constraints are :
• cost constraint ( unit cost from machine ‘i’ to ‘j’ must not increase above Rs 20)

• distance constraint (distance between the departments must not increase by 40feet)

• , Non negativity restriction


SLIDE NO

05/17/2021 RAME 2021 WORK SHOP


18MN72 15
OPTIMIZATION USING MATLAB

Genetic algorithm SA cost function SA distance function


SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL
18MN72 ENGG. 16
ASSUMPTIONS MADE IN THE SIMULATION
MODEL
• The production process line is never starved.

• All process times for operations include insignificant


breakdowns.

• The set up times are not considered, since in a real system


the setup process is usually made at the end of the
working time.

• Three hours of working time does not include breaks.

• Transportation of raw materials is performed by AGV


which is assigned for operations.

• No maintenance process is performed during the working


period.
SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL ENGG. 17
RESULTS AND DISCUSSIONS
SUMMARY OF GENETIC ALGORITHM FUNCTION

S.NO SELECTION CROSSOVER BEST FITNESS ITERATIONS


VALUE

1 Roulette Two point 3351.00 51

2 Roulette Single point 3351.11 88

3 Tournament Single point 3351.07 72

4 Tournament Two point 3351.00 57

SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL
18MN72 ENGG. 18
RESULTS AND DISCUSSIONS
SUMMARY OF SIMULATED ANNEALING COST FUNCTION

SNO ANNEALING TEMPERATURE BEST FITNESS ITERATIONS


FUNCTION FUNCTION VALUE

1 Fast annealing Exponential 3351 551

2 Fast annealing Linear 3351 555

3 Boltzmann Exponential 3351 554

4 Boltzmann Linear 3351 542

SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL
18MN72 ENGG. 19
RESULTS AND DISCUSSIONS
SUMMARY OF SIMULATED ANNEALING DISTANCE FUNCTION

SNO ANNEALING TEMPERATURE BEST FITNESS ITERATIONS


FUNCTION FUNCTION VALUE

1 Fast annealing Exponential 3351 518

2 Fast annealing Linear 3351 536

3 Boltzmann Exponential 3351 537

4 Boltzmann Linear 3351 587

SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL
18MN72 ENGG. 20
RESULTS AND DISCUSSIONS
FINAL SUMMARY

SNO ALGORITHM METHOD BEST FITNESS ITERATIONS


VALUE

1 Genetic algorithm Roulette and two 3351 51


point crossover

2 Simulated annealing Boltzmann 3351 542


(cost function) constant and
Linear function

3 Simulated annealing Fast annealing 3351 518


(distance function) and exponential
function

SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL
18MN72 ENGG. 21
RESULTS AND DISCUSSIONS
S.No Optimal locations Algorithm Iteration
1 341256 GA 51
2 231465 SA – cost function 542
3 423156 SA – distance 518
function

SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL
18MN72 ENGG. 22
FINAL ALTERNATIVE LAYOUTS

Genetic algorithm SA cost function SA distance function


SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL
18MN72 ENGG. 23
SIMULATION USING FLEXSIM

Genetic algorithm SA cost function SA distance function

SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL
18MN72 ENGG. 24
SIMULATION USING FLEXSIM

• The travel distance is 5130 feet. The speed of the Automated Guided Vehicle is 2.5 feet/second. Therefore, the
material handling time when this layout is used will be 34.17 minutes.
• The respective time taken when these two layouts are used will be 37 minutes and 40 minutes. Summary of
these results are given in the table below
SNO LAYOUT DISTANCE ( FEET) MATERIAL HANDLING TIME
( MINUTES)

1 Existing 6526 43

2 Genetic algorithm 5130 34

3 SA distance function 5548 37

4 SA cost function 5986 40

SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL
18MN72 ENGG. 25
CONCLUSION
• It is evident from the results found through MATLAB that the least value of the fitness function can be obtained
only using Genetic algorithm with roulette selection and two point cross over method.

SA cost function SA cost function

SA distance function SA distance function

Genetic algorithm Genetic algorithm

Existing Existing

0 5 10 15 20 25 30 35 40 45 50 0 1000 2000 3000 4000 5000 6000 7000


Distance (feet)
Material handling time (minutes)

SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL
18MN72 ENGG. 26
CONCLUSION
From the bar graph, the layout obtained through genetic algorithm produces least distance and material handling time than any
other.

Reduction of 1396 feet in distance covered and 9 minutes of material handling time from the existing layout is proved in this
methodology i.e, both the material handling time and travelling distance was reduced up to 21% which reduces the overall
production time from 153 minutes to 140 minutes.

SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL
18MN72 ENGG. 27
KAIZEN PROPOSAL SHEET

SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL
18MN72 ENGG. 28
FUTURE SCOPE
• This algorithm can be combined with fuzzy logic and AHP to adopt decision making methods which
involves optimization with the utilization rate.

• Integrating these constraints as well as some parameters can be included to validate its
effectiveness through the application to real cases.

• Other than material handling aspect, objectives like minimizing lead time/work in process may be
considered for the evaluation of each type of layout.

SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL
18MN72 ENGG. 29 29
REFERENCES
• Nordin, N. N., & Lee, L. S. (2016). Heuristics and metaheuristics approaches for facility layout problems: a survey. Pertanika Journal of
Scholarly Research Reviews, 2(3).
• Sharma, P., Phanden, R. K., & Singhal, S. (2013). A Comparative Analysis of Facility Layout Design and Optimization Techniques. In
Proceedings of the 2nd National Conference on Advancements in the Era of Multi Disciplinary Systems (AEMDS-2013).
• Ojaghi, Y., Khademi, A., Yusof, N. M., Renani, N. G., & bin Syed Hassan, S. A. H. (2015). Production layout optimization for small and medium
scale food industry. Procedia Cirp, 26, 247-251.

• Deshpande, V., Patil, N. D., Baviskar, V., & Gandhi, J. (2016). Plant layout optimization using CRAFT and ALDEP methodology. Productivity
Journal by National Productivity Council, 57(1), 32-42.

• Ene, S., Küçükoğlu, İ., Aksoy, A., & Öztürk, N. (2016). A genetic algorithm for minimizing energy consumption in warehouses. Energy, 114, 973-
980.

• Besbes, M., Zolghadri, M., Affonso, R. C., Masmoudi, F., & Haddar, M. (2020). A methodology for solving facility layout problem considering
barriers: genetic algorithm coupled with A* search. Journal of Intelligent Manufacturing, 31(3), 615-640.
• Krajcovic, M., & Hancinsky, V. (2015). Production layout planning using genetic algorithms. Communications-Scientific letters of the University
of Zilina, 17(3), 72-77.
• El-Baz, M. A. (2004). A genetic algorithm for facility layout problems of different manufacturing environments. Computers & Industrial
Engineering, 47(2-3), 233-246.
• Leno, I. J., Sankar, S. S., & Ponnambalam, S. G. (2016). An elitist strategy genetic algorithm using simulated annealing algorithm as local
search for facility layout design. The International Journal of Advanced Manufacturing Technology, 84(5-8), 787-799.
• Lin, Z., & Yingjie, Z. (2019, April). Solving the facility layout problem with genetic algorithm. In 2019 IEEE 6th International Conference on
Industrial Engineering and Applications (ICIEA) (pp. 164-168). IEEE.
SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL
18MN72 ENGG. 30 30
THANK YOU

SLIDE NO

05/17/2021 RAME
18MN72- M.E 2021 WORK SHOP
INDUSTRIAL
18MN72 ENGG. 31

You might also like