Workload and stress are variable and unpredictably changing due to the dynamics of the military environment where the large amount of information available may overwhelm human operators thus leading to poor quality decisions. Phase II of the ANTICO project will develop techniques to support the operator in cognitively compliant ways. In particular, the project extends the development of a general purpose task network queuing estimator of mental workload. Additionally, Markov decision process-based methods for intent recognition, and text/video summarization and task simplification methods will be leveraged to develop a domain independent approach to aiding operators and teams in complex time-stressed environments. These methods are designed to be generally applicable to computer/communications based tasks that are characterized by vast amounts of information and tight decision deadlines. Sample implementations will be developed through a regime of iterative testing of applications in the emergency response domain and military domain(s) to be chosen in consultation with the sponsor.
Keywords: Anytime Cognition, Intelligent Agents, Adaptive Aiding, Mental Workload, Intent Recognition