Fed Panel Asks Today: Why Waymo?

The U.S. Department of Transportation (USDOT) is holding the first meeting today of a new advisory committee focused, in its own words: “on automation across a number of modes.” The committee, made up of an array of experts from a variety of fields, is “to immediately begin work on some of the most pressing and relevant matters facing transportation today, including the development and deployment of automated vehicles, and determining the needs of the Department as it continues with its relevant research, policy, and regulations.”

The big question facing the panel (members listed below) is: Why? Why are we pursuing automated driving?

It was the U.S. government that got us into the automated driving business with the DARPA (Defense Advanced Research Projects Agency) Challenge in 2004. The intent of the original project was to make one third of ground military forces autonomous by 2015. By the time of the multiple Gulf Wars the objective was to develop robotic vehicles that might help protect military personnel from roadside improvised explosive devices.

DARPA’s efforts were preceded by decades of work around the world to perfect automated driving. DARPA was the first organization to identify a commercially viable use case – though we still do not have an automated battlefield.

Google picked up the mantle years later with its own self-driving car effort – advancing the technology philosophically by suggesting that the steering wheel and brake and accelerator pedals be removed and by testing the vehicles on public roads. Google was the first to propose that self-driving vehicles belonged on public roads.

More provocatively, Google presented the case that nothing less than testing on public roads would be sufficient to enable automated driving. In spite of or in reaction to that proposition California regulators required the inclusion of steering wheels and brake and accelerator pedals and the presence of a driver in autonomous test vehicles. It was at this point and on these terms that the current struggle has unfolded.

The business case for Google was to provide transportation options for economically or physically disadvantaged populations. One branch of the Google team ultimately left to pursue commercial vehicle automated driving opportunities (Otto). More than a dozen other startups have targeted public and private driverless shuttle transportation opportunities.

What remains is defining the objective for automated driving and a reasonable and regulated path to market. Auto makers are building Potemkin villages for the purposes of testing cars in real-world-ish scenarios, while technology companies are building simulators to more rapidly compile and analyze the billions of miles necessary to achieve certifiable results. It is clear, though, that nothing less than testing in real world scenarios will advance the technology to the goal of full automation.

But why is the industry suddenly obsessed with this activity – especially given the reality that self-driving cars likely represent a completely different usage scenario not likely to include actual ownership? The interest of the public revolves around testing on public roads and the ultimate objective of reducing highway fatalities. The interest of car makers and regulators is the evolutionary path to autonomy that is expected to make cars safer to drive … or ride in.
Waymo talks about creating a “better driver” which is too high a bar for short-term commercial opportunities. What is a better driver, anyway?

Actually building an automated driving system capable of surpassing human drivers in all driving circumstances will be a decades-long exercise. It is not a reasonable objective.

Toyota Research Institute CEO Dr. Gil Pratt describes this “better driver” dream as the “chauffeur” mode, where the vehicle is doing all of the driving vs. the guardian angel mode where the vehicle assists in what it perceives as urgently dangerous circumstances or simpler tasks. Regulators have done their best to bring guardian angel technologies to the market such as electronic stability control (ESC) and anti-lock brakes (ABS).

Now the National Highway Traffic Safety Administration (NHTSA) is pushing for the adoption of Automatic Emergency Brakes on a voluntary basis. But this voluntary effort points up the limitations of the agency. Were NHTSA to press for a mandate, the process would like require nearly a decade of regulatory action.

It was DARPA that opened up the autonomous vehicle conversation. It was Google that issued the call to action – forcing a regulatory response. Now it is the auto industry, insurance companies, city/state/Federal regulators that are coming together to find common cause.

The “why” of autonomous vehicles boils down to:

[LIST=1]