[2K4-GS-10] AI application: Manufacturing

[2K4-GS-10-01] Development of a Large-Scale Automotive Assembly Work Assignment Optimization Method with Sequential Constraints Using Quantum Annealing

〇Takeshi Moriya¹, Kinya Okada¹, Katsue Saita¹, Yuuji Takahashi², Hiroki Furuichi¹, Satoshi Yoshimura³ (1. Nissan Motor Co., Ltd., 2. Nissan Motor Kyushu Co., Ltd., 3. Groovenauts, Inc.)

Keywords:Automotive Assembly, Combinatorial Optimization, Quantum Annealing, Operations Reserch

Automobile assembly line design involves job allocating work to assembly workers, which is a difficult task. Assembly has many restrictions, such as facility dependencies and specific work order requirements. Additionally, there is a need to minimize unnecessary movement and evenly distribute the workload for multiple stages on the same line. However, assigning 150 tasks to 20 people is a complex combinatorial optimization problem. Traditional mathematical methods are time-consuming due to the problem’s complexity. To address
this, the paper explores the use of quantum annealing for large-scale and complicated optimization problems. However, conventional quantum annealing could not formulate constraints related to part assembly order. Therefore, this paper proposes a new method that combines quantum annealing with sequence adjustment logic to optimize work allocation in the assembly process. As a result of verifying the proposed method by the actual work design, it was confirmed that the work assignment which holds could be obtained in a short time.