PRISA: Infrastructure LiDAR Framework for Proactive Intersection Safety Assessment

arXiv·medium signal

Tam Bang, Hussam Abubakr and Emiliano de la Garza Villarreal (arXiv 2607.16156, cs.LG/eess.SY) place LiDAR on the infrastructure rather than the vehicle to assess intersection risk proactively — scoring near-misses and conflict geometry before crashes happen, rather than analyzing them after. Urban intersections are among the highest-risk points in road networks and vulnerable road users are poorly covered by vehicle-mounted sensing. Notable as an argument for infrastructure-side perception as a complement to autonomous vehicle stacks.