Toward a New Definition of Driving Automation

Not many people have an idea of what the light duty vehicles sold today are capable of. It isn’t necessarily their fault; they’ve been left behind by the way automakers have decided to sell their vehicles.

Try to answer the following questions without using any search engines:

• What is the difference between Level 2 and Level 3 automated driving?
• What is the difference between General Motors’ already deployed Super Cruise and its now delayed Ultra Cruise?
• What is a Level 2+ vehicle?

If you can answer these questions, congratulations! You are likely a member of SAE International, the body that defined automated vehicles in the way that the National Highway Traffic Safety Administration now uses in the US. If you don’t know the answers to those questions, you’re probably a normal consumer, and you might wonder if the confusing classifications and vague definitions are by design.

When Guidehouse Insights published its Sensors and Compute Platforms for Automated Vehicles report last month, one of the goals was to redefine the way we talk about levels of driving automation. The line between advanced driver assistance systems (ADAS) and automated driving systems (ADS) is blurry at best, which leads to additional confusion. If automation really is to become a regular part of the consumer driving experience, there should be a way to quickly and succinctly inform a consumer of what they can expect. Instead of SAE’s Levels 0-5, the following categories would enable consumers to quickly learn what their vehicle is capable of and what hardware is enabling that capability:

• Camera ADAS: A vehicle has basic driver assist features enabled by a camera. Blind spot assist is a common feature on these vehicles.
• Camera + Radar: A vehicle has similar technology to Camera ADAS, but with radar that can measure distance and speed more accurately. At this level, emergency braking is possible.
• Camera + Radar (Hands-On): This is the level of ADAS many might be familiar with. These vehicles feature smart cruise control and lane assist technology. A driver has to keep their hands on the wheel, but they can allow the car to take over some driving functions. In some emergency situations, the vehicle will act for the driver.
• Camera + Radar/Multi-Camera (Hands-Off, Eyes-On): This level allows the driver to take their hands off the wheel. They still need to keep paying attention to the road and be watchful of other vehicles, but the car will steer, accelerate, and brake for them. Some ADS that only allow hands-off, eyes-on driving use lidar, but it is not strictly necessary at this level.
• Camera + Radar + Lidar (Hands-Off, Eyes-Off): This level features lidar, allowing a driver to take their hands off the wheel and eyes off the road. Most ADS designs necessitate lidar at this stage for another layer of redundancy. Importantly, drivers’ brains still need to be “on.” They must be ready to take over control of the vehicle should an emergency occur, so sensors might be used to make sure the driver is awake and facing forward.
• Camera + Radar + Lidar (Fully Driverless): In some scenarios, these vehicles allow the driver to turn their brain off from the driving process. They can read a book, take a nap, or play on their Tesla Arcade. The vehicle can handle the road, and if it can’t, it can slow down and stop safely before alerting the driver.

Are these definitions perfect? No. They encourage OEMs to describe cars in terms of the best they can do. Fully driverless on highways but nowhere else? Still a fully driverless car. However, with these definitions, at least a driver immediately knows what’s on a vehicle and what the vehicle is capable of. With that out of the way, more resources can be spent explaining the even more byzantine EV tax credits.