Adaptive User-Guided Multimodal ENhanced Troubleshooter (AUGMENT) Navy SBIR FY2006.2 Sol No.: Navy SBIR FY2006.2 Topic No.: N06-108 Topic Title: Adaptive User-Guided Multimodal ENhanced Troubleshooter (AUGMENT) Proposal No.: N062-108-0387 Firm: Charles River Analytics Inc. 625 Mount Auburn Street Cambridge, Massachusetts 02138-4555 Contact: Ashley Nunes Phone: (617) 491-3474 Web Site: www.cra.com Abstract: The effectiveness of weapons platforms employed by the Marine Corps is dependent on a maintenance technician's ability to rapidly diagnose and solve problems associated with them. Given that the distribution of novice versus expert maintainers to resolve such problems is unequal, it would be ideal if intelligent agents could be used to provide junior maintainers with troubleshooting assistance for the purposes of enhancing overall development. Given the lack of software tools that specifically address this need, we propose to design and develop the Adaptive User-Guided Multimodal ENhanced Troubleshooter (AUGMENT) to significantly improve the troubleshooting capabilities of junior maintainers through the use of adaptive intelligent agents. Three core components characterize our approach. First, we will perform a work domain analysis on a specific scenario to identify the information requirements associated with problem diagnosis and problem solving amongst maintenance technicians. Second, we will design and demonstrate a tool which uses intelligent agent and multimodal display technology to provide troubleshooting and training assistance to the operator under a variety of conditions. Third, we will conduct initial evaluations to assess the effectiveness of AUGMENT for significantly improving the trouble shooting capability of junior maintainers. Benefits: We see considerable promise in the commercial application of adaptive intelligent agents and multimodal display technologies, especially in domains where large volumes of information must be processed for diagnosis and problem solving purposes by operators of a low skill set. Intelligent agents that are capable of training novices to perform at the level of expert with little or no human oversight have the potential to dramatically improve performance in a number of commercial areas, including financial forecasting and trend analysis, business intelligence, and financial transaction monitoring. In addition, we plan to transition the intelligent agent and multimodal display tool developed for the AUGMENT system into the next version of our Intelligent Agent Toolkit (IAT) product. Return