A fast finite loading algorithm for job oriented scheduling

Job oriented scheduling (JOS) has been the most commonly used technique in actual job shop settings. It loads jobs one by one onto machines, in which all the operations of a job are scheduled in succession. This article presents a JOS model and formulates a fast finite loading procedure, which is the key algorithm of computer-based JOS systems. The algorithm assigns feasible schedule start and finish times to the operations of a job by loading them forward or backward onto the capacity constrained parallel machines. A particular technique is the maintenance of blocks of consecutively scheduled operations on each machine, thereby reducing the search time for finding a feasible time slot for each operation on the corresponding machine. Experimental testing shows that the algorithm has significant computational performance for large size problems.