Raising the Bar on Autonomous Vehicle Safety

completely autonomous vehicles of the not-too-distant future promise great advances in automotive safety and transportation efficiency. But to deliver on this promise, automotive OEMs must go beyond contemporary levels of vehicle autonomy.

Making that leap will require overcoming a unique set of challenges to test automotive radar sensors in advanced driver assistance systems (ADAS) and autonomous driving systems, as well as developing new methodologies for training algorithms that conventional solutions can’t address.

SAE International (formerly the Society of Automotive Engineers) defines six levels of vehicle autonomy, where Level 0 represents fully manual and Level 5 represents fully autonomous.

Today’s most advanced autonomous vehicle systems qualify only as Level 3, meaning they are capable of making some decisions, such as speeding up or braking, without human intervention. Getting from level 3 to level 5 will require many advances, including bridging the gap between software simulation and road testing, and training ADAS and autonomous driving algorithms in real-world conditions.

Software simulation plays an important role in the development of autonomous vehicles. Simulation of environments through software can help validate the capabilities of ADAS and autonomous driving systems. But the simulation can’t fully replicate real-world driving conditions or the potential for imperfect sensor response, something fully autonomous vehicles will inevitably have to deal with.

OEMs rely on road tests to validate ADAS and autonomous driving systems before they go to market. While road testing is and will continue to be a vital and necessary component of the development process, it is time consuming, expensive and difficult to replicate specifically in the area of ​​environmental monitoring. Relying solely on road tests to develop vehicles reliable enough to drive safely on urban and rural roads 100 percent of the time would take decades. For development to occur in a realistic time frame, training algorithms are needed.

Validation of radar-based autonomous driving algorithms is a crucial task. The sensors capture information about road and traffic conditions and send that information to processors and algorithms that allow them to make decisions about how the vehicle should respond in any given situation. Without proper training, autonomous vehicles could make decisions that undermine the safety of the driver, passenger or pedestrian.

Just as people become better drivers with time and experience, autonomous driving systems improve their ability to deal with real-world driving conditions with time and training. And achieving Level 5 autonomy will require complex systems that exceed the skills of the best human drivers.

Premature road tests of unproven ADAS and autonomous driving systems also create risks. OEMs need the ability to Enable validation of real sensors, electronic control unit code, artificial intelligence, and more.

Some systems do not provide a true approximation to real-world driving scenarios. They have a limited field of view and cannot resolve objects at distances less than 4 meters. Some of these systems use multiple radar target simulators, each of which presents point targets to radar sensors and emulates horizontal and vertical positions by mechanical movement of the antennas. This mechanical automation reduces the total testing time. Other solutions create an antenna wall with only a few target simulants, allowing an object to appear anywhere in the scene, but not at the same time. In a static or near-static environment, this approach allows for a test with a handful of targets moving laterally at speeds limited by the speed of the robotic arms.

Image courtesy of Thomas Goetzl, Keysight TechnologiesThomas Goetzl1.JPG

Thomas Goetzel of Keysight Technologies.

Current simulators can emulate a maximum of only 32 objects, including vehicles, infrastructure, pedestrians, obstacles, and other objects. These are far fewer objects than a vehicle traveling down the road can encounter at any given time. Testing radar sensors against a limited number of objects offers an incomplete view of driving scenarios and masks real-world complexity.

Thomas Goetzl is Vice President and General Manager of Automotive and Energy Solutions (AES) in Keysight’s Electronic Industrial Solutions Group, as well as General Manager of Keysight Technologies in Germany. Tom was most recently AES Business Manager for Power Applications, focusing on test solutions for Smart Grid, Power Semiconductors, Batteries and EV/HEV. During his time with HP, Agilent Technologies, and now Keysight Technologies, Tom managed R&D projects, developed sales channels for remote server management products, and was named Marketing Manager in 2005. In 2012, Tom became Product Manager. worldwide marketing of Electronic Test Division. (ETD).

Leave a Comment