Multi-agent system for manpower modeling

Manpower planning is a critical component required to successfully support the operations of an organization. Decisions regarding manpower often involve the estimation of future manpower supply and demand as well as the development of manpower policies involving promotion, attrition, training, and recruitment. Manpower planning is difficult due to the complexities of the interactions between the different manpower elements (i.e. recruitment and training). Manpower modeling uses stochastic techniques to simulate a system and investigate how recruitment, training, promotion, retention, attrition, or any combination of these elements affect the expected level of manpower needed to support the system. Current models are unable to effectively model the complexities of manpower within an organization. Multi-agent systems (MAS) have recently become an important asset in investigating and modeling resource management for organizations. Using MAS as a manpower modeling tool allows for the many different elements of manpower to be modeled as agents to capture the overall behavior of the system while facilitating agents to negotiate and cooperate. In this paper, we propose a manpower decision support tool based on MAS in order to estimate costs and system performance. Results indicate that MAS can be used to effectively model manpower elements for investigating manpower policies.