PENGEMBANGAN ALGORITMA METAHEURISTIK SIMULATED ANNEALING UNTUK PERANCANGAN LINTAS PERAKITAN DENGAN ALTERNATIF URUTAN PERAKITAN DAN PROSES PERAKITAN MENGGUNAKAN KOLABORASI MANUSIA-ROBOT
PT Mattel Indonesia is a toy manufacturing company in Indonesia that has implemented collaborative robot in its assembly line. Implementation of collaborative robot in their assembly line has proven to be more productive and required less manpower. Due to the shorter life cycle of toy products, i...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/67611 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | PT Mattel Indonesia is a toy manufacturing company in Indonesia that has
implemented collaborative robot in its assembly line. Implementation of
collaborative robot in their assembly line has proven to be more productive and
required less manpower. Due to the shorter life cycle of toy products, it is necessary
to redesign the assembly line from time to time. Thus, it is necessary to have an
effective algorithm for assembly line utilizing collaborative robot.
A simulated annealing algorithm has been developed for assembly line design with
alternative assembly sequences (alternative subgraphs) and utilizing collaborative
robot in the manufacturing assembly process. The simulated annealing algorithm
is divided into 2 (two) procedures: outer loop and inner loop. The outer loop is to
run the general procedure of the algorithm, while the inner loop is to develop a
specific procedure in the process of generating new solutions based on process
development results. There are 5 (five) procedures included in the inner loop
system, which are the procedure for maintaining the alternative subgraphs of the
previous solution, the procedure for defining alternative subgraphs, the procedure
for assigning a set of tasks to a group of workstations, the procedure for switching
the tasks, and procedure for exchanging resources.
Experimental analysis of algorithms was carried out using the fractional factorial
design method. Based on this experiment, the algorithm generates feasible solutions
for 9 (nine) experimental data which consist of four optimal solutions related to the
number of tasks 14, 22, 28, and 46, and five near-optimal solutions with the number
of tasks 43, 52, 67, 79, and 92. In addition, the efficiency of the computation time is
57.94% compared with the analytical method. These experiments identified 2 (two)
significant parameters of the proposed algorithm: temperature drop (M) and the
number of iterations in each temperature.
|
|---|