The Defense Advanced Research Projects Agency (DARPA) announced that the team of Charles River Analytics, Draper Laboratory, UC San Diego, and Tempest Technologies has been selected to develop mixed initiative control strategies to address human-automation command and control of large numbers of unmanned air vehicles for the Mixed Initiative Control of Automata-teams (MICA) project. MICA coordinates multi-level control of distributed, semi-autonomous teams of human operators. Its focus is on developing the theory, algorithms, software, modeling and simulation technologies to coordinate multi-level planning, assessment and control of distributed semi-autonomous forces with collective objectives through the hierarchical application of systems and control theoretic methods.
Charles River’s focus will be the development and implementation of a semantic engineering bridge between the operator interface and the underlying hierarchical control system that will allow a single operator to control many vehicles. Building on their evolving intelligent agent technology, Charles River will be applying multiple artificial intelligence techniques such as Bayesian belief networks, fuzzy logic, and expert systems to develop cognitively congruent interfaces that help UAV operators and mission planners understand the limits and strengths of the underlying automation and better facilitate rapid, closed-loop planning in the presence of uncertainty and in the face of adversarial countermeasures.