The Baidu Apollo autonomous vehicle is on display at the 2nd Digital China Summit & Exhibition. [+] The Fuzhou Strait International Convention and Exhibition Center in Fuzhou, Fujian province, China last year. (Photo by Visual China Group via Getty Images/Visual China Group via Getty Images)
Visual China Group via Getty Images
When this technology emerged, the predictions were astounding: one prediction in 2015 predicted that by 2018, autonomous vehicles would be driving around U.S. cities with significant numbers of humans inside.
But as of 2020, AVs have barely caught on in the United States, and other countries, particularly China, are trying to take the lead.
“In the U.S., even the manufacturers are taking a much more cautious, much more calculated view of autonomous driving,” said Jerry Quandt, executive director of the Illinois Autonomous Vehicle Association, “but it’s happening in China, where there are cities where 20 percent of the cars on the road are autonomous.”
It’s not entirely clear where the two countries stand in terms of technological development, although a recent assessment by the South China Morning Post found that American companies have a significant lead in the average number of miles driven without human intervention: 11,000 for American company Waymo, and 1,022 for China’s Pony Eye.
But China, with strong support from the central government, appears poised to dominate the industry.
“China’s ambition is to become a global leader in AI and AVs as part of its Made in China 2025 plan,” the Qvartz analysts wrote. “To achieve this goal, the Chinese government has set a target for 10% of cars sold in China to be fully autonomous by 2030, or roughly 3-4 million vehicles. This has seen China’s AV industry grow from almost nothing a couple of years ago to an exponential growth in companies developing AVs today.”
Experts gathered at the International Smart Cities Symposium in Chicago last week said U.S. growth is slowing for a variety of reasons. I asked Lauren Isaac, who made the 2018 forecast in 2015 while working for the engineering and planning firm Parsons Brinkerhoff.
1 Road hazard ahead
“Most of the OEMs and technology companies are focused on robotaxi deployments, and many of them are promising that Level 5 driverless technology will be available by now,” said Isaac, who now works for Denver-based AV technology company EasyMile.
“Level 5 technology requires fully autonomous driving anywhere, but Level 4 technology requires fully autonomous driving within a limited ‘operational design domain.’ Over the past few years, most companies have realized that developing technology to that level is much more difficult than they imagined, and I think they have changed their stance on aiming for Level 4.”
Why did Level 5 prove to be so difficult?
“I think the main reason is the state of the technology development needed to address the myriad of use cases that Level 5 self-driving cars will encounter, especially at highway speeds. While the technology may be able to handle (hypothetically) 90% of all use cases (e.g. kids chasing a ball in the street, hail storms, construction sites, etc.), there will always be many exceptions. Level 5 is a panacea for self-driving technology, but Level 4 is a more viable option today and for the foreseeable future.”
2. Media Coverage
When one of these unpredictable use cases kills a passenger in a self-driving car, Western media outlets, with their newsworthiness highlighting the newness of the technology, respond with more dramatic coverage than the more than 36,000 fatal accidents that occur each year in human-piloted cars.
“I think the current public opinion, driven by media sensationalism, is definitely another factor in the slowdown,” Quandt said. “And it’s not going to end anytime soon.”
3. Investing in the Status Quo
While self-driving cars are expected to virtually eliminate traffic accidents and fatalities, many industries in the U.S. depend on them for revenue, including law firms, insurance companies and repair companies, Quandt said.
“We’ve built huge businesses around 5 million accidents,” he said. “So how many of those businesses have figured out new business models in this new world we find ourselves in?”
“So that may be another factor. And honestly, the policy is slow in the state legislature because these groups have connections to a lot of legislators and they can block anything if they need to.”
4. Government Organizations
AV manufacturers would benefit from a single policy guidance that could be imposed by China’s central government, whereas in the US auto policy is set by the 50 states.
“State laws are an obstacle,” Quandt said, “because every state has to ratify the policy, so you can’t create a single solution.”
There is also less centralized control over technological development in the US, which Quandt believes will be an advantage in the long run.
“There’s no government driving the solutions. Universities are developing the technology, companies are developing their own technology, and governments – the public sector – are also creating solutions. Their compatibility is just a little slower because there’s no one telling them from above,” he said.
“I actually think that’s better, because I firmly believe that the more people that work on a problem, the better the final solution will be. It won’t be the fastest, but the final solution will be better.”
Speaking at the Smart Cities International Symposium in Chicago, George Crabtree, director of next-generation battery development at Argonne National Laboratory, asked panelists for new predictions.
Crabtree: “We’ve been hearing about self-driving cars for a long time. They will reduce road fatalities and accidents and reduce travel times. We’re not there yet. So what is your prediction for when self-driving cars will become mainstream?”
Quandt: “What do you mean by mainstream? Does that mean a certain percentage of cars on the road will be self-driving?”
Crabtree: “If I had to give a figure, I’d say 20 percent.”
Quandt: “I think in metropolitan areas it will be 20 percent by 2030, 2035. But if you just look at shipping and trucking, I think it will be higher by 2030.”
And the predictions have been consistent: In the U.S., ships and other vehicles are expected to lead the transition, in part because they have better information than most personal drivers, said Matt Stevens-Rich, a program manager at the Electrification Coalition.
“Does anyone know how much they spent on fuel for their vehicle last week, down to the dollar and the pennies?,” Stevens-Rich asked. “That’s a simple question that can be answered by either fleet managers or individual consumers, and that’s part of the reason.”
