A new scheduling algorithm based on approximate reasoning for manufacturing applications

The key to efficiency of manufacturing operations is the ability to schedule various operations involved in the process effectively. Yet, scheduling is an exercise which is not resolvable in unique logical terms for most of the practical situations. Elimination methods, which classify jobs as dominant or front-runners, help to solve the problem using branch-and-bound techniques; however, the subset of the real-life manufacturing problems that satisfy the conditions imposed by the elimination methods is fairly small. To improve upon the range of their applicability, we present a new hierarchical scheduling algorithm based on approximate and heuristic reasoning, resembling the approaches followed by humans. It generates a schedule by tolerating small differences in satisfying inequality conditions on processing times rather than strict evaluation of conditions and by considering those jobs that meet most of the conditions on processing times except one or two as dominant jobs with different degrees of dominance.< >