Automated Vehicles Won't Be a Cure-All

Modern automated driving technology development was initiated by the desire to save lives. Not on the road, actually: on the battlefield. Soldiers in Iraq and Afghanistan were regularly killed or maimed by roadside explosives. The effort began in earnest with the 2003 announcement of the Defense Advanced Research Projects Agency Grand Challenge. Later, the technology and automotive industries saw the potential to cash in while reducing the toll of traffic fatalities. Unfortunately, like most technology advances, automated driving is unlikely to be a cure-all for either business or safety.

DARPA Alums at the Center of Automated Driving

Many of the leading minds of the Grand Challenge went on to form the core of the nascent automated driving business including Sebastian Thrun, Chris Urmson, and Bryan Salesky. Those three all joined the pre-Waymo Google self-driving car project before moving on to ventures of their own. Thanks to carefully staged demonstration videos from Google, its project helped inflate a massive hype bubble.

Hype and High Ambitions

From auto executives to Stanford economist Tony Seba, a vision of zero congestion and zero fatalities was quickly articulated over the past 5 years. Seba went so far as to project that 95% of all passenger miles traveled in the US would come from transportation as a service (TaaS) using automated EVs.

This projection is unlikely to be anywhere near the reality of the situation. Leaders in this sector, including Waymo CEO John Krafcik, are acknowledging the probability that automated vehicles (AVs) will take far longer to spread broadly than many anticipated in recent years. Also, so-called Level 5 automation that can operate anywhere anytime is unlikely before 2030; it may not ever be achievable.

Safety a Major Factor for AVs

On the safety front, automation has been promoted because computers are not susceptible to many of the human flaws that result in decisions leading to 94% of crashes. However, since we are still so far from AVs that can operate beyond a geofenced, low speed, urban environment, it is unlikely that the technology will meaningfully reduce fatalities anytime soon.

According to National Highway Traffic Safety Administration data for 2016, 51% of all road fatalities were in rural areas. The rural fatality rate of 1.96 per 100 million vehicles miles traveled was 2.5 times the rate in urban areas.

Do AVs Work for Rural Areas?

For both technical and economic reasons, these same rural areas are the least likely to be served by AVs in the foreseeable future. Ford executives discussed the business of AVs during a recent demonstration of its automated driving and mobility service development program in Miami. To reach profitability, the services need a degree of scale and utilization.

This requires leveraging the vehicles to generate revenue as much as possible with a minimum of deadheading (time when no passengers or cargo are in the vehicle). Dense urban operations with an intelligent dispatch system can keep deadheading down. Since rural areas have far lower population density, deadheading will be almost unavoidable and dramatically increase service cost. The most vulnerable population, rural residents, are unlikely able to take advantage of AVs and their potential safety benefits until they drive to the urban perimeter to take advantage of first mile/last mile service.

What Does This Mean?

AVs will almost certainly save lives over the long term. But there is almost equal certainty that they won’t eliminate the majority of 1.2 million road fatalities in the next 2 decades. For that we will have to rely on more mundane safety measures like driver training, wearing seat-belts, and more enforcement against impaired driving.