Ensuring Safe Physical AI in Urban Mobility via Hazard-Informed Synthesized Envelopes
2026-08-14 • Robotics
RoboticsArtificial Intelligence
AI summaryⓘ
The authors created a new way to keep robots safe when they move around busy city areas with people. Instead of just setting simple safety rules in one part of the robot’s software, they combined different ideas about hazards from maps, movements, and symbols into one system. This approach helps the robot understand risks better and follow safety rules all the time while working. Their method supports robots moving safely in cities by connecting safety checks directly with how the robot makes decisions and moves.
heterogeneous robotic systemshazard analysisruntime enforcementsafety envelopessymbolic world modelsspatial modelsdynamic modelsphysical AIurban mobilityhuman-robot interaction
Authors
Alexei Odinokov, Rostislav Yavorskiy
Abstract
As heterogeneous robotic systems deploy across diverse urban zones, maintaining safety amid complex human-robot interactions remains a critical challenge. We present a unified framework that bridges systematic hazard analysis and runtime enforcement using hazard-informed safety envelopes. Rather than treating safety as a static constraint isolated within individual software modules, we introduce a cross-layer safety transformation process spanning symbolic, spatial, and dynamic world models. We show how this representation naturally interfaces with physical AI runtime harnesses to guarantee safe urban mobility.