Professional Documents
Culture Documents
net/publication/332430005
CITATION READS
1 246
5 authors, including:
Some of the authors of this publication are also working on these related projects:
A Review on Water Pricing Problem for Sustainable Water Resource View project
Solving assembly line balancing problem using heuristic: A case study of power transformer in electrical industry View project
All content following this page was uploaded by Tan Chan Sin on 19 April 2019.
*
School of Manufacturing Engineering, Universiti Malaysia Perlis, Pauh Putra Main Campus, 02600 Arau, Perlis, Malaysia
3zakimizakaria@unimap.edu.my, 4cstan@unimap.edu.my
2second.author@second.com
Abstract— This project deals with a case study of SALBP in SALBP can be split into two problem types which are SALBP-
power transformer manufacturing. Some workstations in 1 [10] (minimize number of workstations, given cycle time) and
assembly line experienced bottlenecks, while the others suffered SALBP-2 (minimize cycle time, given number of workstation)
with high idle time. Therefore, four heuristic methods are chosen [5], [11], [12].
in order to provide solutions for minimizing number of
workstations and improve the bottlenecks simultaneously. In this There are three types of assembly line [13], which are single
study, LCR, RPW and LPT are successfully minimized number of model [14], batch model [15], [16] and mixed model assembly
workstation from 19 to 16 workstations. Consequently, these line [17], [18]. Besides, there are some assembly line designs
solutions affect the layout. On the contrary, SPT able to improve such as U-shape [2], [19], [20], parallel station [20], [21] straight
the bottleneck by reducing number of workstations from 19 to 17 line [13], [22], loop and rectangular assembly line. This paper
without affect the line layout. Therefore, this has created an option only considers the single model assembly line with straight-line
for engineer to decide which decision should be implement to the design of SALBP-1.
assembly line due to enhance the line efficiency.
The analysis of LCR, RPW and LPT bring up Model I that Workstation Task Task Description
Working Workstation
gives the same result for all three methods. However, the Time (s) Time (s)
Secondary
changes in the task specification of the operator and balancing 1 2 18.53 18.53
winding (sec 1)
result also affect the machine layout. Where the winding Secondary
machine currently at workstation 1 had to be moved to 2 3 26.72 26.72
winding (sec 2)
workstation 6. Whereas the change in task specification involve Secondary
3 4 18.12 18.12
at workstations 6, 7, 8, 9, 10, 14 and 15. Table III shows the winding (sec 3)
change in specification for this model. Secondary
4 5 20.91 20.91
winding (sec 4)
This model only involves small layout changes without 5 6
Secondary outer
22.26 22.26
creating the complicated flow process. In addition, layout wrap
changes model uses 16 workstations and 17 operators in 1 Primary winding 17.95
6 Secondary bobbin 26.59
assembly line compare to current system which consists of 19 8 8.64
soldering
workstations and 19 operators. Primary bobbin
7 11.28
B. Model II: Layout Without Changes Model soldering
7 9 Bobbin assembly 5.50 25.33
This model is a result of SPT heuristic analysis that only
10 Side tape 8.55
involves changes in task specifications on operators. This
model does not change the machine layout and it uses 17 11 E-core insert 18.05
workstations with 17 operators only. Table IV shows the 8 12 Hi-pot test 8.24 28.69
specification of task on the operator for this model. 13 Gluing 2.40
Referring to Table IV, there are changes in the task 14 I-core insert 9.27
specification on operator at workstations 8, 9, 10, 11 and 15. 9 29.10
15 Welding 19.83
The change of assignment task occurred in workstations located
16 Copper foil insert 9.28
at the center of the assembly line, which is at workstation 10 is
proposed to overcome the bottleneck problems. 10 17 Core shield insert 10.55 23.62
TABLE II. COMPARISON OF ANALYSIS BETWEEN HEURISTIC METHODS 18 Bracket insert 3.79
AND CURRENT SYSTEM 11 19 Bracket clamping 13.71 13.71
Heuristic Methods Current 12 20 Functional test 17.39 17.39
Analysis
LCR RPW SPT LPT System 13 21 Varnish 22.50 22.50
Minimum number of
16 16 17 16 19 22 VMI 11.02
workstations, 𝑆𝑚𝑖𝑛
Number of operators, 𝑤 17 17 17 17 19 23 Labelling 4.81
14 26.68
Total idle time, 𝑇𝑖𝑑 Touch-up
96.72 96.72 125.72 96.72 173.72 24 10.85
(sec/unit) soldering
Efficiency line, 𝐸𝑏 (%) 79 79 74.5 79 67
25 Functional test 21.36
Balance delay, 𝑑 (%) 21 21 25.5 21 33 15 23.02
Line balancing loss, L 26 Packing 1.66
21.1 21.1 25.5 21.1 49
(%) 16 27 QC checking 24.11 24.11
Layout changes Yes Yes No Yes -
Change of operator Theoretically, the heuristic method implemented on the
Yes Yes Yes Yes -
specification work
SALBP of case studies succeeded in balancing the assembly
line. The assembly line of 17-P PM should be improved to
Comparison of workstation time between current system,
Model I and Model II
overcome the bottleneck condition and reduce the idling time
between the workstations. These two problems had caused the
Current system Model I Model II imbalanced assembly line and unstable production.
Solutions using heuristic methods such as LCR, RPW, SPT
40 and LPT are aimed to minimize the number of workstations in
35
Workstation time (s)
cycle time, C
30 the assembly line, overcoming the bottleneck condition and the
25
20 idling gap. This step is taken to make ALBP of PM 17-P more
15 efficient and productive. Hence, two proposed models to
10
5 improve the balancing of the assembly line PM 17-P is
0 presented in this paper. In this case, the management team of
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
this electrical industry have some idea to look at the best
Workstation
approach to encounter the problem.
Model I is the solution by changing the position of the
Fig 2 Comparison of workstation time between current system, Model I winding machine and change the task specification on the
and Model II operator. While Model II only involves changes in the task
specification on the operator.
TABLE IV. TASK SPECIFICATION FOR LAYOUT WITHOUT CHANGES Therefore, two models are proposed Model I and Model II
(MODEL II)
to represent the results. The LCR, RPW and LPT are
Workstation Task Task Description
Working Workstation categorized as Model I that involves changes in task
Time (s) Time (s) specification of operators and layout. Meanwhile, the SPT is
1 1 Primary winding 17.95 17.95 categorized as Model II that only involves changes in task
Primary bobbin specification of operators. Both model capable to reduce
2 7 11.28 11.28
soldering
Secondary number of workstations and number of operators. Therefore,
3 2 18.53 18.53 this has created an option for engineer to decide which decision
winding (sec 1)
4 3
Secondary
26.72 26.72
should be implement to the assembly line due to enhance the
winding (sec 2) line efficiency.
Secondary
5 4
winding (sec 3)
18.12 18.12 Hence, for further studies, it is suggested to verify and
Secondary analyze the effects of layout changes. Simulation of changes in
6 5 20.91 20.91
winding (sec 4) layout and task specifications on operators can be proposed to
7 6
Secondary outer
22.26 22.26 recognize the effects of this action.
wrap
Secondary bobbin ACKNOWLEDGMENT
8 8.64
soldering
8 9 Bobbin assembly 5.50 22.69 The authors would like to acknowledge the support from
Universiti Malaysia Perlis (UniMAP) and the Fundamental
10 Side tape 8.55
Research Grant Scheme (FRGS) under a grant number of
11 E-core insert 18.05
FRGS/1/2016/TK03/UNIMAP/03/8 from the Ministry of
9 12 Hi-pot test 8.24 28.69 Higher Education Malaysia and
13 Gluing 2.40 FRGS/1/2016/STG06/UNIMAP/03/6 from the Ministry of
14 I-core insert 9.27 Higher Education Malaysia.
10 29.10
15 Welding 19.83
REFERENCES
16 Copper foil insert 9.28
[1] İ. Baybars, "A survey of exact algorithms for the simple assembly line
11 17 Core shield insert 10.55 23.62 balancing problem", Management Science, vol. 32, no. 8, pp. 909-932,
1986.
18 Bracket insert 3.79
[2] A. Scholl and R. Klein, "Balancing assembly lines effectively – A
12 19 Bracket clamping 13.71 13.71 computational comparison", European Journal of Operational Research,
13 20 Functional test 17.39 17.39 vol. 114, no. 1, pp. 50-58, 1999.
14 21 Varnish 22.50 22.50 [3] K. Fleszar and K. Hindi, "An enumerative heuristic and reduction
methods for the assembly line balancing problem", European Journal of
22 VMI 11.02 Operational Research, vol. 145, no. 3, pp. 606-620, 2003.
15 23 Labelling 4.81 26.68 [4] A. Baykasoğlu and T. Dereli, "Simple and U-type assembly line balancing
Touch-up by using an ant colony based algorithm", Mathematical and
24 10.85 Computational Applications, vol. 14, no. 1, pp. 1-12, 2009.
soldering
25 Functional test 21.36 [5] G. Nicosia, D. Pacciarelli and A. Pacifici, "Optimally balancing assembly
16 23.02 lines with different workstations", Discrete Applied Mathematics, vol.
26 Packing 1.66 118, no. 1-2, pp. 99-113, 2002.
17 27 QC checking 24.11 24.11 [6] O. Battaïa and A. Dolgui, "Reduction approaches for a generalized line
balancing problem", Computers & Operations Research, vol. 39, no. 10,
Fig 2 illustrates the comparison of workstation time between pp. 2337-2345, 2012.
current system, Model I and Model II, where the current system [7] M. Azizoğlu and S. İmat, "Workload smoothing in simple assembly line
balancing", Computers & Operations Research, vol. 89, pp. 51-57, 2018.
consists of 19 workstations. However, number of workstation
[8] D. Giglio, M. Paolucci, A. Roshani and F. Tonelli, "Multi-manned
can be reduced to 16 workstations by using Model I and 17 assembly line balancing problem with skilled workers: A new
workstations by using Model II. mathematical formulation", IFAC-PapersOnLine, vol. 50, no. 1, pp.
1211-1216, 2017.
V. CONCLUSION [9] M. Gansterer and R. Hartl, "One- and two-sided assembly line balancing
problems with real-world constraints", International Journal of
The SALBP is the most common problem faced by Production Research, vol. 56, no. 8, pp. 3025-3042, 2017.
manufacturing companies. The bottleneck condition and the [10] Z. Li, I. Kucukkoc and Z. Zhang, "Branch, bound and remember
idle gap between workstations are two issues that can be seen algorithm for U-shaped assembly line balancing problem", Computers &
in SALBP. In this study, four heuristic methods are addressed Industrial Engineering, vol. 124, pp. 24-35, 2018.
to solve the SALBP of PM 17-P power transformer at electrical [11] S. Karabatı and S. Sayın, "Assembly line balancing in a mixed-model
industry. The heuristic methods able to balance the assembly sequencing environment with synchronous transfers", European Journal
of Operational Research, vol. 149, no. 2, pp. 417-429, 2003.
line and improve the efficiency of the production. However, the
[12] M. Fathi, D. Fontes, M. Urenda Moris and M. Ghobakhloo, "Assembly
results from LCR, RPW and LPT has led to the changes in line balancing problem", Journal of Modelling in Management, vol. 13,
layout and task specification against operators. Different from no. 2, pp. 455-474, 2018.
SPT that reveals a solution, which are affecting to the operator’s [13] M. Groover, Automation, production systems, and computer-integrated
task specification without change the line layout. manufacturing. Upper Saddle River, N.J.: Prentice Hall, 2008.
[14] L. Borba, M. Ritt, and C. Miralles, “Exact and heuristic methods for line balancing with fuzzy processing times,” Journal of Intelligent
solving the robotic assembly line balancing problem,” European Journal Manufacturing, vol. 28, no. 2, pp. 313–336, 2017.
of Operational Research, vol. 270, no. 1, pp. 146–156, 2018. [23] M. Yuan, H. Yu, J. Huang, and A. Ji, “Reconfigurable assembly line
[15] M. Kabir and M. T. Tabucanon, “Batch-model assembly line balancing: balancing for cloud manufacturing,” Journal of Intelligent
A multiattribute decision making approach,” International Journal of Manufacturing, pp. 1–15, Dec. 2018.
Production Economics, vol. 41, no. 1-3, pp. 193–201, 1995. [24] Y. Li and D. Coit, “Priority rules-based algorithmic design on two-sided
[16] S. Qu and Z. Jiang, “A memetic algorithm approach for batch-model assembly line balancing,” Production Engineering, vol. 12, no. 1, pp. 95–
assembly line balancing problem of sub-block in 108, 2017.
shipbuilding,” Proceedings of the Institution of Mechanical Engineers, [25] A. Kalra, S. Marwah, S. Srivastava, and R. Bhatia, “Productivity
Part B: Journal of Engineering Manufacture, vol. 228, no. 10, pp. 1290– improvement in assembly line of automobile industry by reducing cycle
1304, 2014. time of operations,” International Journal of Engineering Research and,
[17] K. Ağpak and S. Zolfaghari, “Mathematical models for parallel two-sided vol. 5, no. 5, pp. 28–31, Mar. 2016.
assembly line balancing problems and extensions,” International Journal [26] M. Hamzas, S. Bareduan, M. Zakaria, S. Ghazali and S. Zairi,
of Production Research, vol. 53, no. 4, pp. 1242–1254, 2014. "Implementation of ranked positional weight method (RPWM) for
[18] L. Tiacci, “Mixed-model U-shaped assembly lines: Balancing and double-sided assembly line balancing problems", In 3rd Electronic and
comparing with straight lines with buffers and parallel Green Materials International Conference, Krabi, Thailand, 2017, pp.
workstations,” Journal of Manufacturing Systems, vol. 45, pp. 286–305, 020183.
2017. [27] K. Park, S. Park, and W. Kim, “A heuristic for an assembly line balancing
[19] M. Li, Q. Tang, Q. Zheng, X. Xia, and C. Floudas, “Rules-based heuristic problem with incompatibility, range, and partial precedence
approach for the U-shaped assembly line balancing problem,” Applied constraints,” Computers & Industrial Engineering, vol. 32, no. 2, pp.
Mathematical Modelling, vol. 48, pp. 423–439, 2017. 321–332, 1997.
[20] I. Kucukkoc and D. Z. Zhang, “Balancing of parallel U-shaped assembly [28] N. T. Lam, L. M. Toi, V. T. T. Tuyen, and D. N. Hien, “Lean line
lines,” Computers & Operations Research, vol. 64, pp. 233–244, 2015. balancing for an electronics assembly line,” Procedia CIRP, vol. 40, pp.
[21] P. McMullen and G. Frazier, "Using simulated annealing to solve a 437–442, 2016.
multiobjective assembly line balancing problem with parallel [29] M. F. M. A. Hamzas, S. A. Bareduan, and M. Z. Zakaria, “A review on
workstations", International Journal of Production Research, vol. 36, no. multi-objective optimization using evolutionary algorithms for two-sided
10, pp. 2717-2741, 1998. assembly line balancing problems,” ARPN J. Eng. Appl. Sci., vol. 11, no.
[22] M. H. Alavidoost, M. H. F. Zarandi, M. Tarimoradi, and Y. Nemati, 18, pp. 10882–10887, 2016.
“Modified genetic algorithm for simple straight and U-shaped assembly [30] R. Schonberger and E. Knod, Operations management. Chicago: Irwin,
1997