A consecutive heuristic algorithm for balancing a mixed - model assembly line type II using a (W-TAWH) model developed for straight and U-shaped layouts

• Published in: IOP conference series. Materials Science and Engineering Vol. 671; no. 1; pp. 12147 - 12160
• Main Authors: Aufy, Samah A., Kassam, AllaEldin H.
• Format: Journal Article
• Language: English
• Published: Bristol IOP Publishing 01.01.2020
• Subjects: Algorithms; Assembly lines; Balancing; Criteria; Cycle time; Desirability function; Global marketing; Heuristic algorithm; Heuristic methods; Layouts; Lower bounds; Manufacturing; Mixed-model assembly line; Performance enhancement; Performance evaluation; Upper bounds; Work stations; Worker; Workstations
• ISSN: 1757-8981; 1757-899X
• DOI: 10.1088/1757-899X/671/1/012147

The development of assembly lines in manufacturing systems has evolved due to consumer expectations and a growing and increasingly competitive global market. The assembly line balancing problem (ALBP) is well recognised in manufacturing and may arise when tasks are assigned to the workstation. In this paper, a problem statement is structured by assigning a task to a worker in a suitable workstation, subject to some constraints, in order to reduce the cycle time and improve performance. This is a new approach, based on worker and task assignment and is thus a worker - task assigned to workstation heuristic model (W-TAWH). It has been developed for straight and U-shaped models. The methodology is described as two stages, the input related data stage and assignment stage. Firstly, the focus is on combining precedence diagram and generating the sequence vector procedure. The second stage comprises development of comprehensive mathematical formulae for the consecutive algorithm to address the straight and U - shaped balancing problem, with minimization or maximation of relevant criteria. This generates performance evaluation criteria according to the number of tasks and workers assigned to the workstation, and these criteria are integrated into a single score using a desirability function approach. Finally, the performance of the developed algorithm is validated through a numerical example. We tested 24 alternative solutions and the fourth solution (1-4-1) achieved the highest score (0.726), which represented the following assignment variables: straight layout, minimum total number of successor tasks rule, and four workers. The performance evaluation is represented by the lower bound of cycle time and upper bound of efficiency.