A System to Observe and Classify Interactions for Assessment and Learning (SOCIAL)

Military personnel are asked to interact effectively, under stress, with individuals from both similar and dramatically different backgrounds and cultures. The desire for automated, operationally deployable capabilities to improve such interactions has created an interest in the development of technologies to sense and report in real-time on the properties of interaction. To address this need, we propose to design and demonstrate a System to Observe and Classify Interactions for Assessment and Learning (SOCIAL). First, we will conduct a requirements analysis to identify a subset of the types of sensor data that are likely to yield insightful information within operational environments. Second, we will collect raw sensor data using only one instrumented individual. Third, we will process sensor data into interaction assessment metrics with an explicit appreciation for individual differences of all participants in the interaction. Fourth, we will enable personnel to review the interaction and assess their actions with respect to individual responses. We will demonstrate the feasibility of this approach within an end-to-end SOCIAL prototype, enabling review of the overall concept by the project team, subject matter experts, and the Sponsor. We will pursue evaluation opportunities for this technology at the National Training Center at Ft. Irwin.