Ridgetop will develop an advanced reasoning-based troubleshooting algorithm to identify the failed components in complex mechanical/electrical systems. The reasoning-based algorithm will be used in an Autonomic Reasoning-based Troubleshooting (ART) management system which will provide support for maintenance personnel in the following areas: (1) the integration with Condition Based Maintenance (CBM) activities; (2) the interpretation of the components interactions; (3) reduction of uncertainty by updating CBM information with in-situ measurement; and (4) an easy-to-use tool with a user-interface and automation support. The significance of the innovation is that expert troubleshooting systems are needed to achieve significant savings by learning what diagnostic actions lead to correct outcomes and minimize wasted time and effort while achieving reduced maintenance costs. Studies show that it is possible to achieve 20% savings on deployed electronic systems. In Phase I, Ridgetop will develop and deliver the design of an ART systems including: - Mapping the fault tree into a reasoning-based troubleshooting algorithm - Interpretation of the component interactions through hierarchical, spatial and temporal information - Reduction of uncertainty with in-situ measurement - Design of an easy-to-use tool including a user-interface and autonomic capability - Validation of the algorithm to quantify the potential benefits through appropriate metrics
Benefit: For todays complex military aircraft, rigorous routine inspection and maintenance procedures are performed to ensure the health of the planes numerous mechanical and electronic systems. While vital, this constant process has seen significant cost increases over the past 10 years as various cost components such as labor, parts, and aircraft downtime rise in conjunction with the increasing complexity of these systems. The proposed autonomic reasoning-based troubleshooting (ART) analysis model can drastically reduce these costs via the rapid identification of faults and failures which typically take significant amounts of time to repair. In addition, Ridgetops innovation will provide an intelligent, user-friendly interface supplying not only the visual diagnosis, but an optimal action plan and rationale to correct the issue as well. Specifically designed for aircraft such as the C-130 and F-15, the innovation can also be applied to the ECSS, F-35, NASA NextGen Shuttles, UAVs, commercial aircraft, and commercial automotive applications.
Keywords: Fault-Tree Analysis, Dynamic Case Based Reasoning, Model Based Reasoning, Bayesian Belief Networks, Built In Test (Bit), Built In Test Equipment (Bite)