Don’t expect to see fully autonomous vehicles in auto showrooms anytime soon. That's according to an informal poll of 300 automotive, energy and technology executives conducted at Bloomberg New Energy Finance’s Future of Mobility Summit last week in East Palo Alto.

Mobility insiders were asked to name the year when U.S. consumers will be able to buy Level 5 fully autonomous vehicles. Nearly 75 percent of respondents to the live poll said the milestone won’t be reached before 2030, and 40 percent said it will come after 2030.

The summit attendees were more pessimistic about the prospects for commercial-scale autonomy than recent automaker announcements suggest. According to BNEF, based on public statements, the consensus among major automakers is that commercial deployment of Level 4 autonomous vehicles will start around 2020; five years later, Level 5 fully autonomous vehicles will likely roll out.

The poll results suggest that the commercial deployment of fully autonomous vehicles is further away than it's depicted in breathless media coverage of self-driving cars -- a view that was shared by a panel of mobility experts. 

“This poll makes me happy because four years ago, even two years ago, this would have been completely different,” said Annie Lien, an independent analyst specializing in autonomous vehicles. “It would have been that everyone thought it would be 2020, or a majority would have thought it was 2020 because people didn’t understand what true Level 5 is, and there were a lot of misconceptions.”

“I’m very glad to see this also,” said Carina Borelius, director of product strategy at Lynk & Co. “It shows that we are realistic. It’s a stepwise approach that we are looking for because safety legislation and attitudes and trust, they need to go forward together; otherwise, we will never, ever achieve this.”

“Automation is going to come,” said Matthew Baldwin, deputy director-general of mobility and transport for the European Commission. “I just don’t think any of us know exactly when or how, or to what extent. Anyone who tells you definitively either has access to different information or isn’t necessarily telling you the whole truth.”

State of play on autonomy

Despite the autonomy reality check delivered at the summit, technology development in the space rolls on. According to an intelligent mobility market update published last week by BNEF, as of this month, 21 states have passed autonomous vehicle (AV) legislation, and 19 of these enable testing of AVs on public roads. Nearly 60 cities globally are hosting pilot programs for AV testing.

Alphabet subsidiary Waymo announced last November it would soon offer ride-hail trips via a fleet of self-driving Chrysler Pacifica minivans to commuters who signed up for its early rider program in Chandler, Arizona. For now, a Waymo employee will ride in the back seat, and passengers can engage a “pull over” button in an emergency. But Waymo’s software will be doing the driving.

Rio Tinto announced in December 2017 it would expand the fleet of autonomous haul trucks at its Pilbara iron ore mines in Western Australia by more than 50 percent by 2019. According to the company, each autonomous truck operated, on average, an additional 1,000 hours at 15 percent lower load and haul-unit cost than conventional haul trucks in 2016.

So it's possible to find autonomous vehicle use cases today, but we're still a long way from mass-market adoption. And if autonomous vehicles are going to play a role in cutting greenhouse gas emissions, we'll need to see the scale-up of electric vehicles, too. 

Manufacturers must earn consumers’ trust

The views expressed at the BNEF summit are an acknowledgement of the work that still needs to be done to bridge the gap between fully autonomous ride-hail pilots and showroom sales.

For Borelius of Lynk & Co, it’s imperative for the industry to convince customers that autonomous cars are safe. “We need to deliver on our promises regarding safety. We need to create a strong brand; people need to be secure," she said." They won’t step into a car if they can’t trust it.”

“The trust element is a huge one,” agreed Gretchen Effgen, vice president of partnerships at nuTonomy. For that to happen, customers must recognize the differences between the five levels of autonomy. For vehicles operating at Level 3 autonomy, she noted, the human driver is supposed to be engaged and ready to take control of the vehicle at all times.

She added, “When there are accidents at that level, but it’s portrayed in the media as an autonomous car having an issue, there’s going to be significant element of trust we’re going to have to deal with as an industry to make that distinction clear.”

Lien warned that as long as humans are behind the wheel, problems will arise. “If the technology works perfectly, consumers are still not going to understand the technology," she said. "They might do something inappropriate, introduce human error. It’s human nature.”

Automakers and technology companies must work together to manage public expectations and educate consumers, she said. “There has been a lack of effort in trying to make sure the consumer or the public in general understands the technology really well and understands the limitations.”

Autonomous deliveries and trucking may lead the way

If mass-market autonomous vehicles are at least a decade away, where should the industry focus deployment in the intervening years? “I think very long term -- maybe 20, 30, 40, 50 years, decades later -- it can be a consumer product,” said Lien. “In between now and then, I think it is going to be smart to focus…[more on] deliveries, transferring goods, and shuttle services or ride-sharing.”

BNEF intelligent mobility analyst Nick Albanese asked panelists in an afternoon session when they thought autonomous vehicles would make up 25 percent of the vehicle fleet.

“I don’t know when they’re going to be a quarter of the vehicles on the road,” said Veronica Siranosian, director of Aecom Ventures. “A more interesting question to think about is: Where would we want those first deployments to be? And how can we invest either in infrastructure or in pilots that utilize connected transit or serve communities that don’t currently have them?”

“It’s hard to say in terms of the urban environments,” said Ryan Popple, president and CEO of electric bus-maker Proterra. “I think we’ll see it in long-haul trucking before anything else. I can see [that] in two, three years, maybe less. Not urban trucking, but from the distribution facility over the road and possibly even split or segregated lanes.”

“I’m more optimistic about AVs when they’re less likely to interact with human beings," he said.