PENGEMBANGAN ALGORITMA VARIABLE NEIGHBORHOOD SEARCH PADA LINTASAN PERAKITAN PRODUK CAMPURAN KOLABORASI MANUSIA DAN ROBOT
The electronics industry in Indonesia recently had positive growth in the first quarter of 2022. This growth is predicted to continue with the increased technology innovation in the products sold. The industry recently applied a collaborative robot (Cobot) for flexibility in responding to market con...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/68358 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | The electronics industry in Indonesia recently had positive growth in the first quarter of 2022. This growth is predicted to continue with the increased technology innovation in the products sold. The industry recently applied a collaborative robot (Cobot) for flexibility in responding to market conditions.
In Indonesia, a company that has already applied cobot on assembly line is PT JVC Electronics Indonesia (JEIN) with operational cost reduced more than 80.000 USD and increased workers’ productivity and safety. As an electronic company, PT JEIN keeps improving to respond to market development by developing various products with same parent depending on consumer demand. This act needs a fast assembly line balancing planning for mixed-model type that utilizes the collaboration of humans and robots as an alternative resource.
This research aims to develop a metaheuristic method Variable Neighborhood Search (VNS) for mixed-model assembly-line balancing problem using a robot, humans, and collaboration of both as alternatives of resources. This research developed the algorithm into six procedures: construction, shaking, local search, operator switching, station allocation, and move. This algorithm was experimented using the full factorial method with influential parameters are number of neighborhood and local search attempts (last one for big data group). Computation result from VNS algorithm reach an optimal solution for three data set and sub optimal solution for other five data with gap of 0,6 – 48%. The computation time for the VNS algorithm is faster than analytics method for all data.
|
|---|