A Comparative Evaluation of Heuristic Line Balancing Techniques

In this paper, we report on a computational experiment designed to assess the efficacy of 26 heuristic decision rules which group work tasks into work stations along an assembly line such that the number of work stations required is minimized. The heuristic decision rules investigated vary from simple list processing procedures that consider a single attribute of each work task for assignment, to procedures which are optimal seeking, but which have had their search terminated through the imposition of a limit on the amount of computation time that can be devoted to each search. Also included are heuristic decision rules which backtrack in an attempt to locate an improved solution, and decision rules which probabilistically search for improved solutions. Our investigation differs from those reported previously, in that the objective in balancing each line is to determine the minimum number of work stations for a given limit on the time available for assembly at each work station (the cycle time). Previous approaches have investigated the problem of determining the minimum cycle time for a given line length. We also compare the results obtained with the optimal solution for a subset of the problems investigated. Both problems which have appeared in the open literature and problems which have been solved for the first time are included. Because a portion of our results differ from those reported previously, we suggest why these differences have occurred. Guidelines are also given to those balancing industrial assembly lines on the choice of the heuristic decision rule to use whether one is attempting to obtain a minimum station balance given a limit on the time available for assembly at each work station, or whether one is attempting to minimize the time devoted to assembly at a work station given a limit on the number of work stations available.