SBIR/STTR Award attributes
Anti-Submarine Warfare (ASW) is increasingly difficult due to modernized threats. Passive ASW detections are ever more challenging in foreign waters as target sets are increasingly quiet and unfamiliar. The Navy has developed improved sensors in the AN/SQQ-89 suite to help operators detect potential threat contacts more quickly and classify them more accurately. Training new operators to use these tools effectively is a critical component of capability development and operational readiness. The Navy has long recognized the importance of training to bring operators’ skills to needed levels, and Metron proposes an innovative approach to advance this priority in the area of ASW training. The fleet has used narrowband signatures as a critical component of ASW classification for years. The Navy benefits from operator training that teaches this approach effectively, and this SBIR provides an opportunity to make strong advances in this area. One difficulty is the dearth of signature examples, environments, and scenarios from which the trainees can learn. Unfortunately, the trainees often see the same data cut multiple times during their training and start to recognize the specifics of particular events instead of the more general threat characteristics represented by those cuts. As junior operators deploy, they can be at a disadvantage in extrapolating their training to new situations. A diverse and agile training environment is necessary to train them to be proficient and adaptive over a variety of targets and environments. Metron offers a two-pronged approach that combines an innovative synthetic-data solution to this data-starvation problem with training methodologies to produce an effective training tool. The tool will be both realistic, resulting in proper skills reinforcement, and inexpensive, yielding wide availability throughout the training community. We will achieve the stated ASW training goals by combining two technologies in an innovative and powerful way. We will (1) augment existing real training data with large amounts of synthetic training data that emphasize passive narrowband signature recognition. This data set combination will produce a significantly richer training set than what is currently available. We will also (2) adaptively display system-like grams to the trainee on the training system in a stimulating game-like fashion to record reactions and evaluate ability. Based on these two components, the system will assess the trainee’s (keystroke and mouse) responses in terms of target prosecution efficacy. Synthetic data that is sufficiently representative of real data enables the system to present trainees with a wide diversity of threat data cuts for training. When combined with training methodologies, this approach should produce training with narrowband data that is more diverse and realistic than the current system for proper skills reinforcement at low cost for increased availability in the training community.
