Software Agent Technology for Large Scale, Real-time Logistics Decision Support

The ability to rapidly adapt mission plans is key to modern operational success. Although this is particularly true for combat forces, it is equally critical for the logistics structure that supports them. Although the availability of logistic decisionsupport systems is increasing, an information overloaded digital battlefield hampers their effective use. This information overload includes both data quantity (large amounts of information of varying type) and quality (information of varying relevanceand usefulness). In addition, the dynamic nature of much of this data further compounds the problem.Software agent technologies may offer some hope of mitigating logistics information overload. Technology Service Corporation (TSC) believes that anend-to-end agent design approach that integrates reasoning models, ontology requirements and communication mechanisms will produce the best results. To this end, TSC proposes to identify, define and demonstrate software agent applicability to logisticsusing its Logistic Site Planning and Operation Tool (LOGSPOT; currently being developed under another Army SBIR) as the baseline system. Since LOGSPOT development will occur in parallel with the proposed effort, the synergy between the efforts should helpensure the straightforward insertion of identified agent technologies during Phase 2.Anticipated benefits of the research include the identification of agent technologies to mitigate much of the up-front data management, collection and synthesis operatoractivities currently part of the LOGSPOT. Because TSC plans to commercialize the LOGSPOT technology, any identified and inserted agent technologies will also be commercialized.