The proliferation of complex autonomous systems has necessitated the development of resilience methodologies for safety, security and reliability of these systems of systems. This work aims to develop methodologies, support algorithms and infrastructure for cross-layer failure detection, diagnosis and correction in intelligent autonomous systems with multiple interacting subsystems acting to meet commands from a supervisory controller. This objective is motivated by the use of intelligent sense-and-control systems in safety-critical roles such as autonomous driving. This work has two key focuses: 
(1) One key focus of this work is cross-layer synergies for system resilience, enhancing the autonomous system's safety through information sharing and coordination between its subsystems. On-line detection and adaptation to failures in actuators, sensors and control or state estimation computation errors are investigated, and methods are proposed which leverage cross-layer subsystem interactions for rapid, on-line failure adaptation.
(2) The second key focus of this work is secure, resilient execution of machine learning (ML) subsystems in roles such as semantic understanding (image classification) and reinforcement learning based control. Compute errors and security threats to ML subsystems during training and inference are detected in real time using reduced-dimension representations of intermediate features within the deep neural networks that make up these subsystems. Suppression of compute errors for safe execution is done through adaptive statistical thresholding of neuron values followed by zeroing of potentially erroneous values.
These two thrusts of work enable resilient intelligent autonomous system design, using bottom-up cross-layer synergies for subsystem failures and module-level resilience methodologies (concurrent error detection, suppression and security modules) in ML subsystems.