Limited-Capacity Supervisors for Discrete-Event Systems

Modern applications seek to develop agents that are capable of performing tasks in real-world environments whose complexity may overshadow the agent's ability to act. These environments require intelligent agents that are capable of achieving their goals with relatively limited capacity. Here we introduce the Limited-Capacity Supervisor Problem, where capacity captures the control and observation capabilities of agents in the system. We formalize the notion of capacity for discrete-event systems, and present a heuristic for reducing a supervisor's capacity usage while enforcing a specification.