The Declarative Framework Approach to Decision Support for Constrained Search Problems

Decision making supported by task-oriented software tools plays a pivotal role in modern enterprises; the commercially available ERP (Enterprise Resource Planning) systems are not able to respond in an interactive on-line/real-time mode. A new generation of DSS (Decision Support System) that enable a fast prototyping of constrained search problems (e.g. layout planning, production routing, batch-sizing, scheduling, transportations, distributions, flow in supply chain, ) is needed. In that context for constrained search problems, the CLP (Constraint Logic Programming) (Apt et al, 2007) techniques allowing declarative representation of a decision making problem provide quite an attractive alternative. Constrained search problems are usually characterized by many types of constraints, which make them unstructured and difficult to solve (NP-complete). Traditional mathematical programming approaches are deficient because their representation of constraints is artificial (using 0-1 variables). This paper discusses declarative framework for decision support in outbound logistic (Fig.1) of SCM (Supply Chain Management) (Simchi-Levi et al., 2003) (Douglas et al., 2000) as an example of use. It will focus on potential areas and processes where decision support or optimization is reasonable, and it will also introduce the concept and an outline of decision support system structures for the SCM in the form of an additional layer of information. The solutions are developed not to substitute, but to be integrated with SCM and with the broader sense Enterprise Resource Planning (ERP) solutions.