Navigating the Simulation Landscape for Self-Driving Cars

Jim talks with Evan McMullen about the state of self-driving car technology, with a special focus on simulators. They discuss the purpose of simulators, levels of simulation, how the world is modeled, gradually ramping up the complexity of the testing world, Tesla's approach, hardware-in-the-loop testing, Waymo's first-mover advantage, simulating the availability of a human intervener, driverless solutions vs driver aid, simultaneous localization and mapping (SLAM), the question of which theories of ethics to use, international standards for functional safety, a liability shield equilibrium, tool-sharing between companies, open source simulators, NVIDIA's DRIVE Sim & other players, standards for interoperability, incentives for cooperation between companies, hardware accuracy, edge case generation, evaluating current offerings for consumers, vibrational tactile feedback vs heads-up displays, when we'll be able to read a book in self-driving car, and much more. Episode Transcript JRS EP 94 - Shahin Farshchi on Self-Driving Tech JRS EP 124 - Jim Hackett on Ford, Electric Cars & More JRS EP 221 - George Hotz on Open-Source Driving Assistance Evan McMullen is a mechatronics engineer at dSPACE, a leading provider of hardware and software for simulation tools into the auto industry.