I tumbled out of bed early one day this week and made it to one of those issues-and-eggs type of breakfasts, this piece of policy jawboning offered by the Urban Land Institute on the topic of autonomous vehicles.
Three experts and a moderator worked through the implications and prospects for that particular transportation innovation, and they had some intriguing thoughts that I’ll share with you here. But with a few days to digest it all, I found myself coming around to a Steve Jobs aphorism, circa 1997.
“A lot of times, people don’t know what they want until you show it to them,” the late Apple founder said to Businessweek magazine.
But what if you show it to them, repeatedly and over a period of several years, and still they don’t appear to want it all that much? Like, in this case, cars that drive themselves. Especially if each of those cars will cost tens of thousands of dollars more than the billions of human-operated vehicles plying the earth’s surface at this moment.
Yes, I know. Almost no such cars are on the market now, at least not the kind where autopilot is the only option, so perhaps people just don’t know yet.
But Tesla vehicles can be put in full self-driving mode — two people have died in accidents when their Teslas were in autonomous operation — and there are public road experiments going on. And several major car manufacturers are talking about bringing self-driving cars to the market within two or three years. For the man and woman on the street, though, autonomous vehicles as a viable purchasing option — or even as a ride-for-hire possibility — are still mostly out there in the mist somewhere.
At the same time, we’ve all been hearing about this coming Big Thing for roughly a decade now. Google researchers jumped into it in a big way in 2009, and a spinoff company, Waymo, has self-driving taxis on the ground in Arizona. Uber and Toyota had test vehicles working in Arizona, too, until an Uber vehicle killed a pedestrian in March.
But all that noise, research and business interest has generated almost no buzz among the public. Do you know anyone on pins and needles waiting for the opportunity to buy, rent or ride in a self-driving car, someone who has brought it up in casual conversation? I don’t, certainly.
Contrast that with ride-hailing, which went from what-is-it to gotta-have-it more or less overnight when Uber and Lyft moved into Austin in 2014. Or those electric rental scooters, which launched unannounced here in April and instantly enjoyed brisk business.
In each of those cases, we didn’t know we wanted an alternative to taxis, or a jaunty and cheap way to move around Central Austin, until someone showed it to us.
The future of autonomous cars, in most of the discussions I’ve heard, has been framed as purely a technological issue. Give the engineers enough time, money and freedom to perfect them, and then an inevitable transition from manual operation of cars to robot operation will occur.
Count me as skeptical. Still.
The Urban Land Institute is made up of architects, engineers, government officials and others in the business of developing cities, and the point of the gathering Wednesday was to discuss how to deal with autonomous vehicles when they come, not if they will. Robert Spillar, the city of Austin’s transportation director, engineer Ginger Goodin with the Texas A&M Transportation Institute and Luis Santi-Merayo, a design director with architectural giant Gensler, spent about an hour kicking around the subject.
Among their insights:
- Goodin, who has researched the subject for the past seven years and seen optimistic technological predictions miss the mark time and again, said it will be at least 10 years until autonomous vehicles are truly ready to operate in all conditions, all the time, on all roads. “I probably tend to be on the pessimistic side,” she said.
- Spillar, pointing to a coming on-road trial in Frisco, just north of Dallas, said he foresees other similarly controlled vehicles being deployed in greater numbers. The Frisco venture was announced in May by the city and its tech partner, California-based Drive.ai. Starting next month, the company’s orange minivans, plastered with “self-driving vehicle” stickers from tip to tail, will begin running a fixed route through an office park and commercial area. Boarding on the vans will occur only at a handful of preset stops, and there will be no charge for a ride. The vans will run on wide streets within an area with little pedestrian traffic at this point.
- Santi-Merayo said that just a few years ago, developers of large buildings or other significant projects had little interest in self-driving cars or what they might mean for their projects. “Now, when I start a meeting, they ask about that from the beginning,” he said. “Right now is the time to start planning for it.”
- That planning, he said, could include parking structures that would have much slimmer parking spaces. After all, when a car drops off its occupants at the building’s front door and then parks itself, there’s no one left to open the doors. So that extra room we all need now to avoid banging the car next to us with our door? Not needed anymore.
- What really flummoxes transportation planners, though, is that transition period of undeterminable length when human-driven and automated vehicles are sharing the road. “How do we manage this mixed system?” Spillar asked.
- Increased safety has always been the primary justification for autonomous vehicles. About 90 percent of all wrecks, Goodin said, are a result of human error. Despite the two deaths this year in crashes involving autonomous vehicles, most analysts think in the end the robots will do a better job of avoiding other vehicles, cyclists and pedestrians. Self-driving evangelists also have predicted a congestion-relief effect on the theory that cars will be able to tailgate and thus squeeze more vehicles into the same amount of pavement.
But don’t count on less gridlock, the panelists said. Taking away the need to drive the car, allowing the passenger to sleep or work on a commute, could actually lead to more miles of vehicle travel. And for autonomous taxi systems, there would be a lot of “deadhead” driving, when vehicles are on the way to pick up passengers.