Autonomous Decision Support for Unmanned Vehicle Control in a Multi-vehicle, Multi-domain Environment Navy SBIR FY2012.2 Sol No.: Navy SBIR FY2012.2 Topic No.: N122-124 Topic Title: Autonomous Decision Support for Unmanned Vehicle Control in a Multi-vehicle, Multi-domain Environment Proposal No.: N122-124-0442 Firm: Soar Technology, Inc. 3600 Green Court Suite 600 Ann Arbor, Michigan 48105-2588 Contact: Jacob Crossman Phone: (734) 887-7606 Web Site: www.soartech.com Abstract: SoarTech will research technologies to address a fundamental problem in current unmanned vehicle operations - reduction in the amount of operator attention required to control a team of unmanned vehicles (UVs). We call our approach Lucid, and the Lucid system will monitor the mission and behavior of a team of heterogeneous UVs and 1. automate the presentation of salient information to the user 2. detect and project important events and alert the user to them, 3. provide context-relevant, high-level inputs, making some actions as simple as one click, and 4. assess user C2 effectiveness, helping to automate task distribution based on context. Lucid will enable supervisory control - allowing operators to spend less time interacting with the system and more time doing other important tasks. To achieve these results Lucid will implement computational situation awareness (CSA), and use it as a basis for intelligent UI control, decision support, and high-level command inputs. The core concepts behind Lucid were prototyped and tested in the MAGIC International Robot Competition where SoarTech teamed with the University of Michigan to win in 2010. Lucid will build on that ground-breaking work by incorporating the naval domain, CSA, C2 effectiveness estimation, and high level command into this framework. Benefits: Lucid's target market is the US military and in particular the warfighter need to control multiple, heterogeneous unmanned vehicles (UVs) in complex missions. Lucid's primary benefit to the warfighter will be as follows: 1. operator interaction time with individual UVs will be reduced 2. operators will be able to neglect all or parts of the UV team for periods of time without losing the aspects of situation awareness that are essential to decision making 3. operator teams will be able to share control of the team, with Lucid aiding them in recognizing periods of high workload and helping to automate task sharing Additionally, Lucid will be designed to tie into existing UV control frameworks such as the Common Control System, allowing operators to seamlessly shift to lower level control when necessary. The final outcome is that in low and moderate complexity situations, fewer operators will be needed to control a heterogenous robot team. In high complexity situations, Lucid will dynamically help to balance workload and ensure that all important situations are addressed by the operators. Return