It was almost six years ago that I harrumphed against the exuberant predictions about the imminent autonomous vehicle takeover of our roads. Today, quietly and without hubris, I can sneak a smile and say I told you so.
A phalanx of tech visionaries have hailed the arrival of “robotaxi” services in the United States and China in the past few months. Even so, I will repeat: fully automated vehicles capable of driving on our open roads will not be here in most of our lifetimes, and certainly not in the lifetimes of old fogeys like me.
Chris Urmson, then the leading driverless engineer for Google’s project, now branded Waymo, predicted in 2015 that they were just a couple of years away from mass deployment, with plans to order 82,000 vehicles. Uber was projecting 100,000 driverless taxis on the roads by 2020. Lyft said most of its rides would be autonomous by 2021.
In 2016, Elon Musk predicted “complete autonomy” for his Tesla cars “in less than two years”. Just last month, a starry-eyed Musk insisted “full self-driving” technology would be solved this year. But as the Financial Times’ Patrick McGee said last July: “The closer they got to a consumer-facing product, the more complex the problem was understood to be.”
To be fair, there has been some progress. GM’s Cruise has won a licence to offer taxi services in an area of northwest San Francisco, but only for employees and between the hours of 11pm and 5am. Google’s Waymo has operated driverless ride-hailing in suburban Phoenix, Arizona, for about two years.
In China, where there is similar excitement about autonomous transport, Baidu’s Apollo Go is operating limited services in the Shougang Park area in Beijing and pilot services in Shanghai, Guangzhou, Changsha and Cangzhou. Alibaba’s AutoX has operated some services in Shenzhen since last January. Pony.ai, in partnership with Toyota, is operating on the roads of a suburb of Shenzhen.
Didi Chuxing is road testing driverless shuttles in several Chinese cities. Truck.tech is developing autonomous truck services in Beijing and Tianjin, Plus and Iveco have been doing pilots on autonomous heavy-duty truck services and SenseTime is introducing autonomous buses in Shanghai.
Geely has announced plans to introduce Zeekr robotaxis in 2024. JD.com is operating autonomous delivery vans and KFC and Meituan are starting to deploy autonomous food delivery trucks.
But the roadblocks to an autonomous takeover remain high, and they do not sit with technology but rather are elsewhere. First is price. The latest-generation Waymo has 29 cameras, radar and lidar, or light detection and ranging lasers.
Such comprehensive sophistication carries a price tag of around US$200,000 per vehicle. This makes autonomous vehicles too expensive for personal use and puts a ride-hailing business model in acute jeopardy.
“Today’s driverless car groups are burning through untold amounts of cash,” McGee noted. Cruise alone has raised US$10 billion and opened a US$5 billion credit line to build more vehicles. Waymo has had to raise more than US$5 billion in the past two years.
Both Cruise and Waymo are valued at around US$30 billion. These are eye-wateringly high valuations and testament to the reality that their survival so far rests not on the viability of their products but on the deep pockets of their parent companies. It should be no surprise that China’s leading autonomous vehicle players are the country’s deep-pocketed hi-tech giants such as Baidu and JD.com.
The second roadblock is the challenge of a normal street environment, with bicycles, motorcycles, kids playing close to the kerb, snow and ice or rainy, uneven roads. As one expert noted: “Electronic chauffeurs will take us wherever we want, in complete safety – so long as we do not need to make any left turns across traffic.”
Third is the limited roll-out of the 5G internet services in most countries. This makes safe autonomous operations impossible in many areas.
Unsurprisingly, most successful pilot programmes have been in well-controlled driving environments – university campuses, science parks, well-monitored highway environments in suburbs and segregated bus- and truck-only lanes. Regulators with an anxious eye on safety have slowed roll-out to a snail’s pace, even though the argument is autonomous vehicles suffer fewer traffic accidents and thus cut road deaths.
This is in part because of the asymmetry of blame, which even Musk acknowledged after the US National Highway Traffic Safety Administration announced an inquiry into 11 Tesla car crashes. He said autonomous vehicle-makers will not be rewarded for the lives they save but can expect to be punished for the lives lost when an autonomous vehicle has an accident.
Beyond the regulators, insurers are also poised for battle on who should assume responsibility when an accident involves an autonomous vehicle. Vehicle manufacturers are in the early days of discovering the scale of future motor insurance liabilities and whether these will threaten commercial viability.
A further threat to the autonomous vehicle business model arises from the faster-than-expected deployment of advanced driver-assistance systems, which are becoming widely deployed in regular, human-driven vehicles. Technologies to manage things such as cruise control, emergency braking and lane-keeping might not replace the human driver, but they increase the safety and reduce the stress of modern-day driving.
My bet is that the most substantial breakthroughs in the next few years will be in controlled areas such as ports or industrial estates and for bus and truck operators using segregated lanes. I am not surprised that companies such as Nuro in the US are concentrating on home deliveries for the likes of Domino’s and Chipotle or Zoox are focused on last-mile delivery for Amazon.
The idea that private drivers will be taking our own autonomous vehicle out for a spin any time soon – an idea that had many leading motor manufacturers spooked a decade ago – is for the time being likely to remain firmly buried in the fertile imaginations of visionaries with spare billions to burn.
David Dodwell researches and writes about global, regional and Hong Kong challenges from a Hong Kong point of view