Modeling Distributed Team Resource Allocation within A Geographical Environment

Decisionmaking functions in many large scale systems are distributed over multiple human decisionmakers who may have access to resources at different locations and are responsible for processing randomly arriving tasks. This leads to a complex team resource allocation and coordination problem. This paper presents a mathematical model and a solution methodology to investigate human team task processing behaviors based on a laboratory experiment. The problem at the current stage of investigation is formulated as a centralized scheduling problem involving multiple tasks and multiple platforms (vehicles equipped with resources). The geographical factors (e.g., locations and velocities) and combination of platforms for task processing are considered. A solution methodology is developed by combining decision tree and the concept of "attractiveness measure." This method is then applied to solve the experimental cases, and the solution methodology generates results similar to those obtained from the human experiment and captures human decisionmaking behaviors. Further model-dats comparison reveals several hidden tendencies in human team behavior, e.g., human teams adapts to tempo by speeding up resource usage but sacrificing efficiency. This model and solution methodology will be extending to a distributed case to resemble the coordination between team members, and a normative-descriptive model will be developed to analyze human decisionmaking.