A cognitive architecture for emergency response

Plan recognition, cognitive workload estimation and human assistance have been extensively studied in the AI and human factors communities, resulting in many techniques being applied to domains of various levels of realism. These techniques have seldom been integrated and evaluated as complete systems. In this paper, we report on the development of an assistant agent architecture that integrates plan recognition, current and future user information needs, workload estimation and adaptive information presentation to aid an emergency response manager in making high quality decisions under time stress, while avoiding cognitive overload. We describe the main components of a full implementation of this architecture as well as a simulation developed to evaluate the system. Our evaluation consists of simulating various possible executions of the emergency response plans used in the real world and measuring the expected time taken by an unaided human user, as well as one that receives information assistance from our system. In the experimental condition of agent assistance, we also examine the effects of different error rates in the agent’s estimation of user’s stat or information needs.