Planning under Uncertainty

In a supply network numerous random events occur that may have significant implications for the performance of the value-adding processes in the complete supply network. These events may be machine breakdowns, increased transportation times or random demand variation. Random demands and the resulting forecast errors are presumably the most dominant factor. Unfortunately, in addition to external random influences, inadequate planning approaches that neglect capacity constraints during the planning process, such as the MRP (materials requirements planning ) concept may add random variation to the development of the value-adding processes and make its outcome unpredictable.

Planning approaches used in Advanced Planning Systems (and more generally in Operations Management) can be devided into the following categories:

  • Planning approaches that emphasize the dynamic nature of the planning problem and which assume that all data are known with certainty. These planning concepts, e.g. lot sizing models under dynamic conditions and scarce capacity, are discussed here.
  • Inventory management approaches that emphasize the randomness of the data. These approaches typically use prociples from stochastic inventory theory, were the demand is modelled as a stationary process. More details about these planning approaches, that usually neglect capacities are here.
  • The best fit to the real problem is provided by planning approaches that simultaneously consider randomness under dynamic conditions. These are dynamic planning approaches, that take into account the randomness of the demand (e.g. dynamic stochastic lotsizing models).

More information on issues related to uncertainty is here: