September 2021
USRA and NASA scientists have been awarded Center Innovation Fund to develop High-Performance Quantum-Classical Hybrid Model for Flight-Operations Anomaly Detection. NASA Ames Research Center has awarded NASA - USRA scientist team a Center Innovation Fund (CIF) award to develop a high-performance scalable and explainable Machine Learning model for flight-operations anomaly detection. The core model will be a novel classical deep-learning model, with a design that can be complemented by quantum computing, which has demonstrated the capacity to improve and accelerate learning through statistical sampling from energy-based distributions. The project will introduce multiple innovations including semi-supervised learning and explainable feature engineering. It will focus on the analysis of flight-operations data as a use case. However, the applicability of fast, accurate, and scalable anomaly detection goes far beyond flight operations as the proposed high-performance model will be application agnostic and generalize to other use cases. For example, the model can be integrated with diagnostics and fault-management systems in the early-formulation phase of new space missions, such as DAVINCI+.
The project is a collaboration between the Data Science and Quantum Group in the Intelligent Systems Division at NASA Ames Research Center. Team collaborators include: Dr. Thomas Templin (Ph.D.) (NASA) PI, Dr. Milad Memarzadeh (Ph.D.) Co-I (USRA), and Dr. Ata Akbari Asanjan (Ph.D.) (USRA) and the Quantum Artificial Intelligence Laboratory (Dr. Eleanor Rieffel (Ph.D.) Co-PI (NASA), Dr. Salvatore Mandrà (Ph.D.) (KBR Wyle), and Dr. Aaron Lott (Ph.D.) (USRA).