Methods for solving setup planning problems

Abstract

Automatic setup planning has been an active area of research for a long time. Traditional approaches of setup planning have been using decision tree, decision table, group technology, algorithms and graphs. A number of experts systems also have been developed for setup planning. Expert systems follow forward chaining or backward chaining. The forward chaining supports data-driven reasoning; whilst the backward chaining supports goal-driven strategy. A number of soft computing tools have also been used in setup planning. Prominent among them are fuzzy set, neural networks and evolutionary optimization. Fuzzy set theory takes care of uncertainty and imprecision in the information. Artificial neural network can learn from the experience. Evolutionary optimization techniques help to provide optimum or near optimum solution. The goal may minimize the number of setups, manufacturing cost or resource consumption. The popular evolutionary optimization methods are genetic algorithms, particle swarm optimization and ant colony optimization. Nowadays cloud computing is also finding its application in setup planning.