Model Integrated Computing and Autonomous Negotiating Teams for Autonomous LogisticsThe MICANTS project seeks to develop efficient negotiation protocols for distributed problem solving in the logistics domain. Using the multi-agent systems paradigm, we propose to develop software systems composed of light-weight components that, through cooperative negotiation, carry out complex, distributed problem solving activities. The objective of this highly proactive, autonomous logistics system is the maximization of lead-times for maintenance planning, parts procurement, stocking, and positioning of machinery to be maintained (such as aircraft). Furthermore, due to the complexity of such systems, we claim that a system-integration technology is needed which supports the rapid integration of complex components. Drawing from previous research and experience in model-integrated computing, we will develop modeling formalisms designed for the logistics domain that allow the capture of any and all relevant information that is needed to synthesize and integrate a system. The resulting models will be used by system generators to build executable instances of the modeled system. The use of MIC allows the domain-specific design phase to be separated from the system implementation phase of a project. By designing and modifying aspects of the system in a domain-specific, graphical language, we reduce the cycle time for implementing modifications to a system when the domain-aware generator component does not have to be updated to accommodate the modifications. The expertise of how to translate domain-specific models into executable systems is contained wholly in these generators, which [presently] must initially be built by highly-skilled computer specialists.
|