Technologies for Enabling Warfighter Intuitive Decision Making Navy SBIR FY2012.2 Sol No.: Navy SBIR FY2012.2 Topic No.: N122-144 Topic Title: Technologies for Enabling Warfighter Intuitive Decision Making Proposal No.: N122-144-0443 Firm: Icosystem Corporation 10 Fawcett St. Cambridge, Massachusetts 02138 Contact: Eric Bonabeau Phone: (617) 520-1020 Web Site: www.icosystem.com Abstract: Traditional concepts of neural-machine interface use brain signals as: control inputs; to assess neurocognitive states for modification of information delivery to improve human performance; to guide selection of image sets. We propose a revolutionary concept of interfacing Electrical Geodesics' technology (sensors and brain signal analysis) with Icosystem's interactive evolutionary computation (IEC) (search and decision support) for certain data mining and discovery tasks. IEC combines the unique ability of humans to instantly detect patterns with the computer's ability to generate millions of solutions. This approach enables a significantly broader exploration of model space while always satisfying the model constraints. While the task described is simple for a proof-of-concept, the implications can dramatically improve the quality and thoroughness for a range of analytic and discovery tasks. Benefits: In the era of "Big Data", this technology will have broad application in military as well as commercial settings in which large quantities of information must be quickly and accurately analyzed for effective decision making in high-risk and high-stress operational settings. In the military, systems that help warfighters effectively process this information are urgently needed, especially in view of personnel reductions at a time of increasing information flow. Additionally, in both the government and commercial sectors, a form of "high-throughput screening by humans" is required to explore vast amounts of data that have been seen or collected before; indeed, no machine on its own will discover patterns without being trained on them. The proposed approach leverages and amplifies limited human resources to make high-throughput screening by humans possible. In our experience, applications can be found beyond real-time situations: for example, new sensor data collected in oilfields, refineries, roads and bridges, cars, etc, can be meaningfully exploited using this approach. Return