Embedded Training Techniques for Target Discrimination Systems

Complex systems such as the AN/SPS-74(V) Periscope Detection Radar automatic target recognition system presents difficult training challenges for enhancing and maintaining operator proficiency. Technical challenges arise from the nature of software automation. The most difficult recognition cases can usually be trained effectively only with actual sensor data. Keeping the proficiency training current is also a challenge. Automation algorithms evolve in time as improvements are developed. System features such as display format and tools also evolve. Hence, the training must evolve at the same pace. 3 Phoenix, Inc. proposes an innovative approach to develop a very lightweight, portable training delivery software infrastructure system that coexists with the tactical software and leverages the tactical software functions as much as possible. This eliminates the need to duplicate or emulate the operational environment on a training platform. The proposed embedded trainer includes a novel Contact Evaluation Assistant function which aids the operator in understanding the decision making processes of the automation software, thus improving the collaboration between operator and automation. The proposed approach will be an innovative training solution that is applicable to any system that implements automated target recognition with operator collaboration.