Sponsor:

AFOSR

Research Team:

Heather L. Machtmann, Edward A. Pohl, Scott J. Mason, Manuel D. Rossetti

Universities Involved:

University of Arkansas

Start Date:

01/01/05

End Date:

12/31/06

Summary:

This project has the objective of working toward information superiority through enhanced cognitive decision support, integration of course of action analysis, knowledge management and mining, and pattern recognition and learning into the network and improve the logistician’s ability to convert data and information into relevant knowledge and understanding. Also, through network and modeling analysis. this project attempts on creating a framework that can be used to construct mathematical and logical models of sense and respond (S&R) military logistics networks.

Information architectures for network-centric logistics rely heavily on automation for achieving the goals of adaptive, rapid response to support operations. S&RL as envisioned by the OSD Office of Force Transformation, emphasizes total asset visibility, automated planning through agents, and flexible supply and transportation to achieve these goals. What is missing in this architecture is the role of the human actor, or team of actors, at the edge of the logistics infosphere. The logisticians who will operate the S&RL network must collaborate from multiple locations, and they will need carefully crafted tools and procedures to do their work effectively. For example, human interaction with agent societies and other networked artifacts present special challenges for coherent decision making and coordinated action. The classic problem of evaluating the proper assignment of functions between human beings and machines have special resonance in the S&RL context because of the complexity of the emerging net-centric planning problem, the requirement to support rapid re-planning, and the uncertainty of logistics needs in the dynamic modern battlefield.
It is well understood that automated technologies that fail to inform users of agent actions and system state increase risk, particularly in novel, unpredictable situations (Billings, 1996; Woods, Tittle, Feil, & Roessler, 2004). While many logistics activities may be routine and predictable, the S&RL of interest for this research are not. Logistics planning in the context of war must be very flexible. Plans must be adapted quickly and in response to unexpected events. This generally happens in the context of distributed teams facing many coordination challenges including geographical distance, disparate time zones, and the transmission of sensitive and classified information. Successful S&RL depends not only on effective agent technologies, but also on smooth coordination between agent technologies and human teams.
*New modeling paradigms for supply chain simulation
*New simulation methodologies, mathematical models and techniques for the optimization, performance evaluation, and improvement of future military logistics support networks
*New methods of modeling and incorporating human performance issues and collaborative decision making within an S&RL environment