Using Complexity for Manpower Modeling: Feasibility Study

We examined whether the class of complexity-based model(s) known as Agent-Based Models (ABMs) could be a useful decision-support tool for personnel planning and management. In using ABM, a system is modeled as a collection of autonomous, decision-making agents. ABMs are built using an object-oriented programming language. Each agent, the agent’s environment, and the schedule that controls the model run are independent objects that can be matched in a variety of ways. A major strength of ABM is its ability to simulate real interactions between individuals and groups allowing for a wide variety of feedback, adaptation, and negotiation behaviors. However, ABM’s results are sensitive to initial conditions, and the reliability of such results is limited to ranges of outcomes linked to ranges of input parameters. In examining a variety of ABM applications—including biological, behavioral, and organizational—we determined that ABMs have dealt with the kinds of issues important to the Navy and that, while not a perfect analogy, an ABM supply chain type of model would meet many of the Navy’s personnel modeling requirements. Given the possible benefits of using an ABM, we feel that there would be value in building a prototype “proof-of-concept” ABM to test its utility.