To address the NASA LaRC need for innovative methods and tools for the diagnosis of aircraft faults and failures, Physical Optics Corporation (POC) proposes to develop a new Active Integrated Diagnosis with Ensembles (AIDE) system, based on Bayesian network modeling, ensemble learning, and context-aware reasoning. This approach incorporates an active fault diagnosis system architecture, a block-level Bayesian-network-based context model, and a context-aware reasoning and severity assessment engine, which enable us to meet NASA aviation safety mission requirements for reliable and accurate diagnosis and assessment of adverse events with minimal uncertainty. The system offers constantly updated aircraft health context, which guides the active queries on aircraft health management systems to minimize the uncertainty along its progress path in the context model and make statistical inference and diagnosis, providing rank-ordered lists of diagnoses, severity assessments, and uncertainty measurements. In Phase I, POC will demonstrate the feasibility of active diagnosis of aircraft faults and failures by establishing context models and building and testing a preliminary prototype, which will demonstrate TRL-2 by the end of Phase I. Integration and validation issues will be explored through communication and collaboration with manufacturers. In Phase II, POC plans to develop a fully functional prototype, including software and supporting hardware, and demonstrate its fault diagnosis capability on a family of adverse events in the AirSTAR testbed. The results demonstrated will offer NASA the capabilities to diagnose and assess adverse events and improve aviation safety.