Assembly line balancing with equipment requirement and parallel workers: an heuristic algorithm

Purpose Balancing assembly line means to assign assembly tasks to several stations, respecting precedence constraints so as to complete the final product with a minimum expense of resources and costs. Aim of this paper is to suggest an heuristic algorithm to solve the balancing problem considering three important real aspects. First, each task, or elementary assembly operation, requires specific equipments that differ from task to task involving cost and space requirements. Second, large products (e.g. cars, automatic machines or white goods), can be handled by more than one worker contemporarily processing different tasks in the same station (parallel workers). And third, parallel workers operating in the same station, can generate space or handling conflicts, considering task assembly position. Design/methodology/approach An heuristic algorithm is presented to design assembly lines considering a set of configuration parameters: task ordering priority rule, maximum number of parallel workers per station, maximum worker idle time allowed and maximum number of equipments per worker: varying the value of each parameter is possible to determine different assembly line configurations. A Visual Basic application is developed to evaluate such configurations with respect to four significant key performance indexes: equipment requirements, employed workers, workers saturation, number of assembly stations. The best solution can be chosen considering both plant priorities and total costs. Originality/value The literature proposes several models that deal with the assembly line balancing problem, nevertheless few of them investigate real world systems. This paper proposes a balancing heuristic algorithm considering the distinctive features of real assembly lines, e.g. equipment requirement. This significant characteristic is usually ignored by the literature, neglecting the equipment duplication. Practitioners and plant managers can benefit from this algorithm exploiting its distinctive feature: the generation of several assembly line configurations. The different solutions can be compared using the key performance indexes assessed by the algorithm. The analyst can select the final configuration considering most relevant assembly line features with respect to the specific plant conditions.