The development of complex aviation weapons systems involves flight testing at various stages of the design cycle. Test pilots are highly trained individuals with flight and assessment skills necessary to evaluate the performance of test articles. Test pilots might subconsciously use their high level of expertise and skill to compensate for adverse test article traits. What is needed is a system that allows for automatic quantification and documentation of pilot workload expenditure and performance for use during the flight test cycle. This will provide test pilots with enhanced means to identify adverse test article traits early in the design-test cycle. We propose to develop a self-contained, flight hardened system that predicts pilot procedural behavior and quantifies the experienced cognitive loading using fused sources of real-time information derived from a normative behavior model, human operator, aircraft state, and situational context. We call this the Automated Human System Performance Assessment (AHSPA) tool. The AHSPA tool will increase the probability of detection of adverse system traits during the flight test phase and thereby reduce/avoid costly fixes in fielded systems. We have a signed letter of intent from the USN Test Pilot school to work with our team as a potential transition customer.
Benefit: The proposed work will benefit the government in several possible ways as follows: 1). Reduced design-test cycle cost by identifying adverse test article traits early in the design-test cycle, 2). Academic support in USNTPS program for instruction on human workload characterization methods, 3). In direct testing support with programs of record. The AHSPA tool uses real-time measures of workload and a model that predicts how much workload a pilot should devote to a particular mission task element (MTE). The prediction model in the AHSPA tool can do this as a function of pilot experience (expert versus novice). As long as the workload trace provided by the AHSPA tool lies within the bandwidth of expected workload, the test task has most likely been performed adequately. Deviations of the actual workload from the predicted workload are likely indicative of problems with the test article that need to be further studied. AHSPA allows for deep drill down into causal data during after action review (ARR). The cost for eliminating adverse system traits in fielded systems can take on catastrophic proportions. The AHSPA tool represents a means to identify adverse test article traits early in the design-test program. This capability will result in considerable cost risk reduction in complex weapons systems development cycles. As we have arranged for collaboration of the USN Test Pilot School (USNTPS) as a possible transition customer, we feel that the AHSPA tool could be used in programs tested at USNTPS and in academic instruction of modern methods of flight test human performance assessment. Our proposed method and approach was well received at the USNTPS. Several potential applications and transition programs were identified. Among potential platforms are the P-8, the new maritime patrol aircraft, which is in the midst of testing, the CH-53K, the Marines' new heavy-lift helicopter, which will be tested within the next five years, and a number of unmanned aerial systems (UAS) platforms. One potential focus in the P-8 test program will be on quantifying performance of the individual 9 crew members who use electronic workstations for anti-submarine warfare, anti-surface warfare, intelligence, surveillance and reconnaissance in broad-area, maritime and littoral operations. This means that the AHSPA tool could be developed for deployment on each crew member as part of a test program to quantify real-time workload throughout a typical P-8 mission profile. Applications of the AHSPA to tool in the CH-53K test program may be focused on the evaluation of avionics displays that are specifically designed to reduce or eliminate the adverse effects of brownout.
Keywords: pilot workload assessment, pilot workload assessment, data fusion, cognitive and procedural behaviors, MicroCore processor miniaturization, human state characterization
