Intel’s framework for AV safety gains global acceptance
Autonomous Vehicle Technology announces 2020 ACES Award Winner in Autonomy | Collaboration category
With safety being one of the biggest roadblocks to AV adoption, the auto industry has become increasingly vocal about the need for measurable, technology-neutral standards for AV safety. Demonstrable assurances are needed so governments know when it’s time to grant AVs the license to drive. Support is gaining quickly for Intel’s RSS (Responsibility-Sensitive Safety) model, first introduced in 2017 by Mobileye, to develop those assurances and standards. RSS formalizes human notions of safe driving into a verifiable model with logically provable rules and defined responses. It is different from the safety decision-making systems in AVs today that are based on probabilistic artificial intelligence. Because of AI’s probabilistic nature, Intel recommends a separate, deterministic layer—or safety envelope—that checks the safety of an AV’s decisions based on well-understood definitions of what it means to drive safely. This model prescribes safety and does not rely on millions of miles of driving history to teach safety. At the 2019 CES, Valeo embraced Intel’s model, Baidu reported the first open-source implementation of it, and China ITS approved a proposal to use RSS as the basis for a forthcoming AV safety standard. Momentum for RSS accelerated in July, when industry leaders put their weight behind a white paper, “Safety First for Automated Driving,” a non-binding organized framework for the development, testing, and validation of safe automated passenger vehicles. It was back by Aptiv, Audi, Baidu, BMW, Continental, Daimler, Fiat Chrysler Automobiles, HERE, Infineon, Intel, and Volkswagen. A shared vision to reduce traffic fatalities through driverless technology has yielded a wide range of approaches throughout the industry. The framework reconciles the many different approaches into a cohesive whole where only the best and safest approach is taken.
Click to view AVT’s 2020 ACES Award Winners best in class of