Combining thermal infrared, 77 GHz radar, and solid-state LiDAR on smart utility poles to guarantee 99.9% detection accuracy.
Dense fog, heavy blizzard snowfall, and blinding sunlight glare can reduce optical camera detection accuracy by more than 80%. When an autonomous vehicle depends solely on its own forward cameras, adverse weather forces drastic speed reductions.
V2XSense installs multi-modal perception clusters on elevated traffic poles. By combining 77 GHz radar (which penetrates fog and snow) with long-wave infrared (LWIR) thermal cameras (which detect human body heat regardless of lighting), the system generates continuous 3D occupancy maps. ``` +-----------------------+ Multi-Modal Sync +-----------------------+ | Radar + LiDAR + LWIR | --------------------------------> | V2XSense Edge Fuser | | Elevated Smart Pole | | Cross-Modal Attention | +-----------------------+ +-----------------------+ | v ETSI CPM Sidelink Broadcast "Pedestrian at Crosswalk" ```
In sub-zero winter trials in Tromsø, Norway, V2XSense detected pedestrians and road debris with 99.4% precision through blinding snowstorms, broadcasting alerts that triggered safe vehicle stops 4 seconds before onboard sensors detected the hazard.