Continual Verification and Assurance of Robotic Systems under Uncertainty
Verifying robotic systems (RS) to obtain assurances that they operate safely and reliably is often very challenging due to the uncertainties associated with measuring their parameters and those of their environments, and with using such measures to detect disruptive changes and to adapt to or recover from them.
This project will develop a theoretical foundation comprising novel run-time methods for RS change-point detection, diagnosis of system-level RS failures, and synthesis of new RS configurations enabling recovery from these failures. We will integrate these methods into an end-to-end methodology for the engineering and assurance of RS operating in uncertain, rapidly changing environments, and validate their feasibility by developing two demonstrators.
The first demonstrator will use research mobile robots to implement a proof-of-concept infrastructure maintenance application. For the second demonstrator, the project will develop a fully-fledged simulator of an unmanned underwater vehicle used for offshore pipe inspection.
Lead Investigator: Dr Radu Calinescu, University of York
For information about the COVE project please contact Dr. Radu Calinescu