2020/05/14 by Eduardo Sánchez Morales, Morales, Eduardo Sánchez, Richard Membarth +13
Engineering · #Autonomous Vehicle Technology and Safety #Distributed #FOS: Computer and information sciences #Parallel #Real-time simulation and control systems #Robotics (cs.RO) #Vehicular Ad Hoc Networks (VANETs) #and Cluster Computing (cs.DC)
paper · pdf · doi:10.48550/arxiv.2005.06773
openalex publication_date 2020/05/14 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Due to the current developments towards autonomous driving and vehicle active\nsafety, there is an increasing necessity for algorithms that are able to\nperform complex criticality predictions in real-time. Being able to process\nmulti-object traffic scenarios aids the implementation of a variety of\nautomotive applications such as driver assistance systems for collision\nprevention and mitigation as well as fall-back systems for autonomous vehicles.\n We present a fully model-based algorithm with a parallelizable architecture.\nThe proposed algorithm can evaluate the criticality of complex, multi-modal\n(vehicles and pedestrians) traffic scenarios by simulating millions of\ntrajectory combinations and detecting collisions between objects. The algorithm\nis able to estimate upcoming criticality at very early stages, demonstrating\nits potential for vehicle safety-systems and autonomous driving applications.\nAn implementation on an embedded system in a test vehicle proves in a\nprototypical manner the compatibility of the algorithm with the hardware\npossibilities of modern cars. For a complex traffic scenario with 11 dynamic\nobjects, more than 86 million pose combinations are evaluated in 21 ms on the\nGPU of a Drive PX~2.\n