NHTSA and Autonomous Vehicles (Part 1): The 5 levels of automation

During the last month, the National Highway Traffic Safety Administration (“NHTSA,” the agency that didn’t redefine “driver” in February) held two public hearings on autonomous vehicles (“AVs”), one in Washington DC on April 8 and another at Stanford, in the heart of Silicon Valley, on April 27.  In keeping with what you might expect, press reports of the two events suggested that the Silicon Valley gathering attracted the voices of people more enthusiastic about the promise of AVs and more intent on urging NHTSA not to let regulations stifle innovation in the field.

These public hearings are an important and positive sign that the NHTSA is serious about moving forward with the regulatory changes that will be necessary before autonomous vehicles become available to the general public.  But before turning to what went down at these hearings (and to buy some time for me to watch through the full video of the second hearing), it’s worth pausing to give some background on NHTSA’s involvement with autonomous vehicles.

NHTSA has shown increasing interest in automation since 2013, when it issued an official policy statement that defined five levels of vehicle automation.

No-Automation (Level 0): The driver is in complete and sole control of the primary vehicle controls – brake, steering, throttle, and motive power – at all times.

In essence, Level 0 means ordinary passenger cars.  This includes every vehicle built during the first century or so (at least: see below) after the invention of the automobile.

Function-specific Automation (Level 1): Automation at this level involves one or more specific control functions. Examples include electronic stability control or pre-charged brakes, where the vehicle automatically assists with braking to enable the driver to regain control of the vehicle or stop faster than possible by acting alone.

In the developed world, at least, the vast majority of new passenger vehicles sold in the past decade fall into this category.

Historical side note: depending on how broadly or narrowly you define an “automated” feature, the introduction of the first “Level 1” vehicles occurred, at the earliest, with introduction of cars with electronic anti-lock brakes and computerized automatic transmissions in the 1970s.  But mechanical automatic transmissions had been introduced decades earlier and mechanical anti-lock brakes had been used in aviation since the 1930s, and electronic ABS and automatic transmissions were not really doing anything different than their mechanical predecessors from the perspective of a human driver.  So arguably, the “Level 1” era did not end in the developed world until cars with active electronic stability control systems started hitting the road around the turn of the millennium.

Combined Function Automation (Level 2): This level involves automation of at least two primary control functions designed to work in unison to relieve the driver of control of those functions. An example of combined functions enabling a Level 2 system is adaptive cruise control in combination with lane centering.

An increasing number of vehicles have Level 2 automation, with Tesla’s Autopilot technology being perhaps the most advanced and best-known example among vehicles sold to the general public.

Limited Self-Driving Automation (Level 3): Vehicles at this level of automation enable the driver to cede full control of all safety-critical functions under certain traffic or environmental conditions and in those conditions to rely heavily on the vehicle to monitor for changes in those conditions requiring transition back to driver control. The driver is expected to be available for occasional control, but with sufficiently comfortable transition time. The Google car is an example of limited self-driving automation.

Despite this last sentence, Google has since made it clear that they are trying to develop a vehicle with…

Full Self-Driving Automation (Level 4): The vehicle is designed to perform all safety-critical driving functions and monitor roadway conditions for an entire trip. Such a design anticipates that the driver will provide destination or navigation input, but is not expected to be available for control at any time during the trip. This includes both occupied and unoccupied vehicles.

This type of fully AV is what most people think of when they hear the term “self-driving car.”  Several companies, from startups to tech giants to multinational automakers, are already working on Level 4 vehicles that are in various stages of testing and development. Some have tried to introduce the term “Level 5” to describe vehicles that are not merely capable of fully autonomous operation, but that are actually incapable of have a human drive the vehicle. It is also worth noting that the SAE has its own six-level rubric for AVs that goes from level 0 to level 5. To avoid confusion, I will stick with the NHTSA rubric and simply refer to “autonomous-only” vehicles as “full” Level 4–i.e., they are always autonomous, rather than autonomous only under some circumstances.

With that background in place, my next post (update: linked here) will focus on the promised benefits of Level 4 automation–or, to be more precise, the drawbacks of not moving quickly toward Level 4 automation.  The final post (update: linked here) will discuss what was said at