A resilience compiler for edge / Physical-AI systems.
Same real video, two systems, same failures. Left: a normal system that needs the cloud. Right: Failsafe. Pick a domain to see it on a dashcam, a drone, or a robot.
Link healthy. Both systems are watching the zone.
Both sides see the same real detections. The left side is a modeled baseline that needs the cloud. On the warehouse scene the right side runs the compiled policy; on the other domains its labels are illustrative. Footage: NVIDIA PhysicalAI-SmartSpaces (CC-BY 4.0), KITTI, VisDrone, JRDB, YCB-Video.
Drawn from the recorded experiments, with full provenance. Nothing here is a mock-up.
Each dot is one real alert. When the cloud is dead (offline), all 57 arrive under the 2 second limit. When it is slow but alive, every one of 281 arrives too late, on the same configuration. Catching slow, not just down, is the whole point, and the fix is a verified on-device mode, not a smarter prompt.
Checked on real video from five domains: a warehouse, a car, a drone, and two robots. The search comparison and the full method are in the report.
Every condition the compiler looked at, one per row. Green passed on every clean run. The others say why there is no verified mode, so nothing is guessed.