Scalable enterprise systems are designed to support collaboration and rapid decision making in response to exogenous stimuli.  We posit that the ability to make tight yet achievable promises in response to requests from consumers or other businesses is a fundamental capability of intra-enterprise and inter-enterprise business automation.  Therefore, this research investigates techniques for real-time promising by discrete build-to-order environments with dynamic order arrivals.  Results of the research are directly applicable to the increasing number of manufacturers that sell built-to-order products direct to customers via the Internet and to a future where collaborative commerce freely occurs among dynamically recombinant business partners.  Phase I of the project strategically focuses on the primitive mechanisms required within a single domain for real-time promising and dynamic collaboration.  Phase II would reuse these primitive mechanisms in the investigation of dynamic collaboration between a network of domains.  (*)

A vast literature exists on scheduling to meet specified due dates.  There exists, to a large extent, a misconception that the problem of setting due dates has basically been solved.  However, very little effective research has been done on due date setting, or promising, which is perhaps the most important operational level decision.  This will be the focus of the research.  The existing techniques for due date promising that are scalable consider order characteristics only in an aggregated fashion.  However, in discrete build-to-order environments the variance of flowtime across all orders is quite large, and algorithms are needed that can appreciate individual order characteristics on a detailed level when computing due dates.

This research project develops algorithms for real-time due date promising that consider current time-phased availability of resources and material, existing commitments, and the current system state.  Various alternates (resources, paths, raw materials, sources, and sub-components) create combinatorial complexity.  To increase performance, a combination of both optimal algorithms with good scalability such as shortest path and computational heuristics are considered.  One of the heuristics to be examined is based on a novel, even controversial, idea:  for the purposes of promising, the time when a resource will be able to process an operation can be estimated with sufficient accuracy by considering only a partially ordered task plan.  Current support for this principle is based on practical experience but little scientific evidence.

Algorithms are implemented in an object-oriented, memory-resident, multi-threaded architecture for detailed study and empirical evaluation.  Joint research occurring at the intersection of Industrial Engineering (IE) and Computer Science (CS) is necessary to develop highly scalable techniques in this project.  Ultimately these techniques will allow computation of a promise date in a fraction of a second for an industrial-sized system.

The proposed approach is novel but has high potential impact and a high probability of success.  The problem definition is nontraditional.  It is anticipated that this catalytic project will reset paradigms about real-time due date promising and create a new thread of research.  Cross-fertilization between research on production systems and database systems will occur.  This project enhances technology transfer in a straightforward manner:  by tackling realistic and important problems.  An outstanding set of industrial partners will keep the research team abreast of rapidly changing practical requirements.  As an added benefit, most of the investigators have extensive practical industry experience in areas relevant to this project.

An interdisciplinary team from IE and CS will perform the research.  Such collaboration is essential for this project, since the algorithmic aspects of promising and advanced computational approaches for technological realization are intertwined when performing research in a large-scale systems context.  Results are of theoretical interest to both fields and will be incorporated into courses offered by both departments.

For problems or questions regarding this web contact database@cs.ou.edu.