Resolving Visual “Edge Cases” Leads to Safer Autonomous Vehicles - White Paper
ANSYS SCADE Vision is a powerful software tool that can address inherent weaknesses in AI-supported perception systems for autonomous vehicles by pinpointing hundreds of potential edge cases quickly and affordably. Edge cases are objects in a traffic scenario that have less recognizable shapes than standard traffic signs and automobiles, such as a baby stroller or a person in a wheelchair. Obviously, misclassifying such objects poses a significant safety risk.
SCADE Vision minimizes the cost of edge case resolution and perception system validation by automatically detecting the weaknesses in these systems. As it reviews video-based sensor data while exercising AI-based perception system algorithms, SCADE Vision augments key areas for closer analysis. Based on that augmentation, which is intended to reproduce the wide variety of conditions that a deployed system will encounter, SCADE Vision automatically identifies and flags any abnormalities, with no human intervention.
Early adopters of this innovative technology have reported a 30X acceleration in detecting edge cases compared to manual methods. Perception engineering teams can achieve a significant time, cost and safety advantage by capitalizing on this advanced technology from ANSYS.