CRAY Colloquium: From Sensors to Driving Rules: End-to-End Analysis for AV Assurances
The computer science colloquium takes place on Mondays from 11:15 a.m. - 12:15 p.m. This week's speaker, Sebastian Elbaum (University of Virginia), will be giving a talk titled "From Sensors to Driving Rules: End-to-End Analysis for AV Assurances"
Abstract
The adoption of autonomous vehicles (AVs) requires rigorous driving assurances. However, specifying and verifying safety properties remains a formidable challenge. This difficulty stems from three primary hurdles. First, the environment that must be explored against driving rules is vast and complex. Second, a semantic gap exists between high-dimensional raw sensor data and the high-level symbolic concepts required for safety reasoning. Finally, the rapid evolution of underlying models, coupled with escalating development costs, necessitates the early detection of violations before they manifest in late-stage deployment. Reflecting on our work through this landscape, we have found that addressing these hurdles requires a family of end-to-end analyses that move beyond traditional validation and verification.
One key insight is that we can bridge the semantic disconnect by lifting raw sensor data into structured scene graphs and flows. Our work shows that by explicitly capturing entities and their relationships, we gain the symbolic abstraction needed to reason about safety at a higher level. Furthermore, by leveraging these abstractions to assess coverage and shifting focus toward runtime monitoring and safe-by-construction training, we have identified significant potential for building inherently safer systems. Finally, as these models evolve, our preliminary results suggest that we must analyze the alignment between internal model reasoning and emerging vehicle behaviors to effectively detect and repair driving rule violations.
Biography
Sebastian Elbaum is the Lowell Professor in the Department of Computer Science at the University of Virginia where he co-leads the Lab for Engineering Safe Software (LESS Lab). He aims to build dependable autonomous systems. He is the recipient of an NSF Career Award, an IBM Innovation Award, a Google Faculty Research Award, Amazon Scholar, an FSE Test of Time Award, five ACM SigSoft Distinguished Paper Awards, and multiple best paper awards. He regularly serves in program committees at the top software engineering and robotic conferences, and has served as Program Co-Chair for ISSTA, ESEM, and ICSE, and as Steering Committee Chair for ICSE. He is an Adjunct Senior Fellow for Emerging Computing Technologies at the Council on Foreign Relations connecting autonomous systems and AI with policy in national security. He is a council member for the CRA Computer Community Consortium. He is an ACM Fellow and an IEEE Fellow.
Category
Start date
Monday, May 4, 2026, 11:15 a.m.
End date
Monday, May 4, 2026, 12:15 p.m.
Location
Keller Hall 3-180