CADR-RAS Webinar: Adversity ‐ and Adversary ‐ Robust Adaptive Positioning Systems with Integrity

Over the past several decades, major research advances in localization (positioning) and navigation technologies have been achieved, enabled in large part by transformative new learning techniques and the increase scale and quality of training data.

But advances in research don't easily translate to deployment on robots, autonomous vehicles, drones or for people, for a multitude of reasons that are both challenging but also tantalizing research opportunities. To be useful, these systems must maintain reliable performance in the face of adversity: degraded lighting or atmospheric conditions, in slowly or rapidly but always changing environments, as well as in adversarial environments where intentional or unintentional jamming or spoofing is present. These capability requirements in turn suggest major areas for targeted research that are largely unexplored, such as the development of introspective localization systems that "know when they don't know", improving their performance and facilitating human handover, oversight and trust. In this talk we'll cover in detail our work on a Centre for Advanced Defence Research in Robotics and Autonomous Systems (CADR-RAS) project in adversity- and adversarially - resilient robot positioning and navigation systems. We will also provide an overview of the broader range of research we're performing in this area and highlight some of the most exciting areas and opportunities for ongoing exploration and collaboration.