© 1968 by British Computer Society
A permutation procedure for job-shop scheduling
Mathematics Branch, A.E.R.E., Harwell, Didcot, UK
A new method is proposed for scheduling jobs through a factory. The jobs consist of a sequence of operations, each operation requiring a number of resources of different types. The objective is to plan the start times of the operations so as to minimise the cost the jobs being late, subject to the sequence constraints being satisfied and the demand for resources not exceeding the supply. The problem is formulated in terms of optimising a permutation, and conditions for a locally optimal permutation are defined. A procedure is described for obtaining such locally optimal permutations and subsequent results show significant improvements over heuristic techniques such as the shortest operation and least slack rules.