Computational complexity of algorthms fo rMRP and JIT production planning problems in enterprise resource planning systems

摘要

An enterprise resource planning system (ERP) is the information backbone of many manufacturing companies. At the core of ERP is a conventional material requirements planning (MRP) production planning system or a variation of MRP when just-in-time JIT) principles are used in manufacturing. MRP and JIT both organize production planning into a hierarchy of long-, medium- and short-range problems. In all there are eight different problems. Some are common to MRP and J IT, others are specialized for a particular system. This paper analyses the computational requirements of these problems. This is important for ERP because it plans for large numbers of products (e.g. 50000 products at SM Company and 44000 products at States Industries in Oregon) at large numbers of locations (e.g. 82 locations in 21 countries for Visteon Automotive Systems of Michigan and 19 locations at Boeing). We show that adequate algorithms exist for some problems, but better algorithms are needed for other problems if ERP is to provide useful production plans.