Nonprofit, nonpartisan journalism. Supported by readers.


Robot cars will change your life — maybe

REUTERS/Stephen Lam
A screen displays views from various onboard sensors in a Google self-driving vehicle before a presentation at the Computer History Museum in May.

It’s easy to chuckle at the thought of robot cars. As a kid growing up, I spent afternoons watching old re-runs of “Knight Rider”: KITT, the electronic Trans Am, talking away with then-hunky David Hasselhoff as they fought off the bad guys. The vision always seemed like a mixed bag to me, spending hours of your life talking to a machine, a robot car taking all the fun out of driving.

But there’s one simple reason that self-driving cars are inevitable: The current status quo is very bad. Because we all do it almost all the time, it’s easy to forget that driving around in cars all the time is extremely deadly. Over the last 50 years, an average of 40,000 people were killed annually by distractible humans driving cars. (That’s the mind-blowing total of 2 million Americans.) Unless you’re a lumberjack or an Alaskan fisherman, getting in your car is the most dangerous thing you do each day.

Frank Douma is a research fellow at the University of Minnesota’s Humphrey School of Public Affairs who recently hosted a conference on autonomous vehicles (aka robot cars) and explained to me that more than 90 percent of auto crashes are the fault of a driver making a mistake.

“After nearly 100 years of having cars, we’ve made them very safe,” Douma told me this week. “The one area we’ve not been able to really improve is the driver interaction with the car. There’s only so many other things we can do.”

David Levinson is an engineering professor at the University of Minnesota who attended this month’s conference. He explained to me that the key for autonomous vehicles is that they can react far more quickly and precisely to their surroundings.

“We could go down from 33,000 to a few hundred deaths per year by car,” Levinson told me. “In mixed environments, speeds can be regulated so that cars go much slower. People might be more wiling to travel at slower speeds in neighborhoods when they don’t have to stop at stupid traffic lights. And we won’t have the option to be more aggressive, like we can right now.”

Unlike many harried urban drivers today, robot cars would always stop for a child crossing the street or give plenty of room to a bicyclist. Saving tens of thousands of lives, while making cities safe again, is an inspiring vision.

Impeccably driven robot cars would also greatly expand our road capacity. Compared to mistake-prone humans, over twice as many robot cars might fit onto a lane of highway, which could make traffic jams (and freeway expansions) obsolete.

City of Minneapolis Public Works Department 2012

What happens to driving?

But in the short term, the driver remains the biggest challenge for robot cars. Completely autonomous cars (known in the industry as “level 5”) are still a long way off. Until then, drivers’ cars would still have some level of responsibility. You wouldn’t be able to go to sleep or drive your robot car drunk, but you might be able to use a computer, read a book, or watch the latest episode of the inevitable “Knight Rider” remake.

And according to Douma, who is a legal scholar, this gray area poses tricky legal questions.

“Personal liability won’t go away,” he told me. “You’ll see quite a bit of interaction and negotiation between manufacturers and the NHTSA (National Highway Traffic Safety Administration) about what is going to be an acceptable level of safety. You probably won’t see so many of those mom-and-pop injury lawyers, but you’d see more on the class-action side.”

Liability is a sticky problem because, outside of drunk driving, car drivers have remained largely immune to responsibility for the tens of thousands of deaths on the road. (A case like Amy Senser’s is the exception that proves the rule.)

But that’s beginning to change as texting and mobile technology uproot the delicate balance of automobile violence. Robot car technology will add a whole new level of complexity to a system, and nobody is quite sure how that might work.

Robot cars reshape the city

At the same time, the ability to daydream while driving is appealing to stressed-out commuters. As Levinson explained to me, robot cars might lead to even more driving than we see today.

“Autonomous cars will be faster on average, and as a result they’ll increase the distance people are willing to travel, “ Levinson told me. “They will also reduce the cognitive burden of drivers, and so people will be willing to spend more time driving. Both things would lead to further suburbanization.”

Courtesy of Brendon Slotterback
Each year we spend hundreds of millions of dollars on expensive parking lots and freeway expansions; what if those massive projects were obsolete in a decade or two?

In a world filled with robot cars, drivers would have less stressful commutes. Picture cities where mega-commutes of 50 or 100 miles were commonplace (for those who could afford the surely more-expensive gas).

And in the city core, on-demand robot car sharing programs might mean that more people wouldn’t own one of their own. Robot cars could park themselves, ending the need for expensive downtown lots or car storage in apartment buildings. (Imagine Hasselhoff going out to dinner and using KITT as a robo-valet.)

Brendan Slotterback is the sustainability program coordinator for the city of Minneapolis, and sees a lot of promise in cities filled with robot cars.

“Theoretically, if you can save space in the right-of-way by moving people more efficiently; you could use that space for something else,” Slotterback said. “More space for non-motorized transportation or green space or whatever. And there’s the parking question: What do you need so many ramps in the downtown area? We could use those for redevelopment areas.”

Theory vs. real life

All this might sound like a fantastical vision, but any one of these outcomes would be a game changer for urban policy. Each year we spend hundreds of millions of dollars on expensive parking lots and freeway expansions; what if those massive projects were obsolete in a decade or two?

There are still a lot of questions surrounding robot cars. How quickly could they be adopted across the country? Could they be “hacked”? How will governments handle frightening privacy concerns? And, here in Minnesota, can they be designed to work in the ice and snow? (Currently, bad weather really ruins the performance of cameras and sensors.)

Personally, I’m still a bit skeptical about whether the technology can really work. Solving the driver liability issue seems devilishly complex. But robot cars are happening right now, and the safety benefits are impossible to ignore. Like it or not, robot cars mean that driving could look different in a decade: thriving walkable cities, far fewer parking problems, and traffic jams melting away in favor of a class of supercomputer supercommuters. In no time, those long-term highway plans might look as silly as Hasselhoff’s hairstyle.

Comments (20)

  1. Submitted by Todd Hintz on 11/14/2014 - 08:37 am.

    Bad Drivers

    Humans are the weakest link in the driving chain.

    This morning I did a little test to see how well drivers knew a few simple rules. There was a gentleman waiting at at a street corner to cross and get to a bus stop. In Hennepin County all corners are considered crosswalks, whether marked or not.

    So I stopped and waited. Three cars from the opposite direction drove past without so much as a pause. The worst though was a red SUV who blazed by on my right, passing close to my car and using the parking lane and bike lane to do so. Both are illegal in the state of Minnesota.

    That’s one huge advantage of self-driving cars: they not only know all the rules of the road, but also follow them. Unfortunately the issue of bad weather blinding sensors is only one of several huge hurdles autonomous cars have to cross. They also can’t navigate around pot holes, a big problem here in our Great Frozen Northland. Also every light and object must be coded into the program before the car knows it’s there. Add in a new light and the car can’t pick it up on its own.

    Plus the program doesn’t know a ball rolling into the street may be followed by a child, so it doesn’t adjust its speed accordingly.

    Self-driving cars are a heck of a good idea, but unless there’s some major breakthrough in the process, they’re still more than a decade or two from full production.

  2. Submitted by Neal Rovick on 11/14/2014 - 08:45 am.

    My big question is whether people will want to relinquish their control on driving.

    In an effort to gain a few extra minutes people speed, go faster than road conditions suggest, run stop signs, run yellow and red lights, tail-gate, don’t let people merge, don’t stop for pedestrians, cut-off other drivers and bicyclists, etc., etc.,

    My guess is that a vehicle that FOLLOWED all laws and courtesies would be hated and regarded as a bigger intrusion than NSA spying or Obamacare.

    And then there is the matter of the box on the road or bag blowing in the wind–would an emergency stop be made for those things? How big an object would it stop for?

    And would it let anyone drive in road conditions we had in the recent winter storm (“no unnecessary travel advised”)?

    • Submitted by Bill Lindeke on 11/14/2014 - 11:24 am.

      The right to be a menace

      A robot car system will be far more efficient and reduce travel times overall. If people want to race around in vehicles, that can be a recreational activity somewhere that it doesn’t kill people and harm the quality of life in our cities.

      It’s a distinct possibility that we’ll look back on our “happy driving” era like we look back on the golden age of cigarettes. Both have killed immense numbers of people.

    • Submitted by Todd Hintz on 11/14/2014 - 11:58 am.


      Fear of this battlestation will keep the systems in line.

      Oh wait–wrong venue.

      When the technology hits the market people will want to adopt it for a variety of reasons.

      -The early adopters will grab it because it’s damn cool!
      -Another segment will pick it up because it makes sense for safety reasons.
      -Eventually it will become the de facto car on the market and you won’t be able to buy an olde fashioned manual car.

      Further down the line, lower insurance rates for the self-driving car will prompt a lot of people to make the switch. There will still be die hards and collectors who hang on to the old cars (I still have a stick shift car), but they’ll increasingly become a smaller segment of the market as time wears on. It may even get to the point where the manual cars are only allowed on certain roads with a low speed limit so they can cause the least amount of damage should something go awry.

      I expect that will take a hundred years to come about though.

      • Submitted by Jonathan Ecklund on 11/14/2014 - 02:22 pm.

        Grand Moff Molnau?

        I wonder if control and regulation of the system would have to fall under a new agency? Would each state regulate the infrastructure for driverless cars under some sort of titular government head? I guess only IF each state’s system was unable to connect to another state’s system. It’s interesting, for sure… though I think I agree with some of the other posters here, that I’d rather see an increase in walkable cities as opposed to an impetus for more sprawl. In other words… i’d rather live on Bespin than Coruscant.

  3. Submitted by Joseph Senkyr on 11/14/2014 - 08:55 am.


    I’m excited about the technology of self-driving cars, but I’ve never been able to understand the common claim that massively increasing VMT, (as no-doubt-single-occupancy vehicles drive in from the burbs, drop off their passenger, then drive back to the outskirts to park for the day before repeating the process eight hours later) is going to be a gain for urbanism or sustainability.

    • Submitted by Gerald Abrahamson on 11/14/2014 - 11:09 am.

      Vehicles are NOT “single use” or “exclusive use”

      The cars are used all day to move people–so they usually do NOT “park”–unless there is no demand for rides at that moment for that particular vehicle. If no demand for a vehicle, then it gets repositioned for the next anticipated rush. Electric–not gas. So no need to go to the gas station–there are automatic recharge stations all over the road network to recharge vehicles as needed. If there is no demand at the moment for all of them, the ones near recharging stations can be recharged while waiting.

      Why single occupancy? Could be–at highest price. The system would know who is going where and when, so vehicles move inward–toward city center from the outskirts. Pick up passengers with a common/close dropoff point (same employer?) while also eliminating unneeded long side trips. People save by sharing the fare–if they choose to share.

    • Submitted by Todd Hintz on 11/14/2014 - 11:40 am.


      My guess is the car would simply find the nearest parking space and wait there for the day, much as they do now. The big change is the car would query a database and see where there’s an open spot rather than endlessly circle the block looking for a spot.

  4. Submitted by Jeff Klein on 11/14/2014 - 09:17 am.

    If they increase commutes, we all lose

    I’d rather we reduce deaths by reducing driving and sprawl. If we end up commuting even more, more people will just die from sedentary lifestyles.

    Also, Prof. Levinson makes an interesting point about how “People might be more wiling to travel at slower speeds in neighborhoods when they don’t have to stop at stupid traffic lights.” We could do this now, without self-driving cars, by replacing traffic lights with yield signs and round-abouts. A street environment where everyone moved continuously at 20 mph instead of 0-35-0-35 would be better for drivers, pedestrians, and bikers alike.

    • Submitted by Bill Lindeke on 11/14/2014 - 11:21 am.

      the trade-off

      I am hopeful that robot cars ceasing to terrorize our potentially walkable cities outweighs whatever happens on the freeways and suburbs. The main point is that it’s hard to imagine a worse outcome than the status quo.

  5. Submitted by Bill Hansen on 11/14/2014 - 09:28 am.

    It can’t happen soon enough…

    Robot cars are zipping all over San Fracisco right now, logging millions of accident free miles.

    Drivers will be eliminated gradually, starting with self-parking, lane sensing and crash avoidance – all being used in the real world now.

    If it can save millions of lives and debilitating injuries, make our cities much more pleasant and drastically reduce costs, simple morality demands that we do it ASAP.

    • Submitted by Todd Hintz on 11/14/2014 - 11:47 am.


      Unfortunately the product isn’t ready yet for prime time, so we can’t roll it out all over the place no matter how wonderful it seems. The car has logged thousands of accident-free miles, but it’s all on the same few hundred miles around Google headquarters. They’ve got that area mapped out and would need to map the rest of the city (and country) for cars to go anywhere else.

  6. Submitted by Frank Phelan on 11/14/2014 - 09:55 am.

    Legal Concerns

    My legal concern is the law of unintended consequences. It’s application in this area will be very interesting indeed. It’s greatest benefits and worst costs are probably beyond our imagination right now.

  7. Submitted by Pat Thompson on 11/14/2014 - 11:55 am.


    Since the cars are networkable (hence Bill’s worry about hacking), they can also share info on things like potholes… which I can’t imagine they wouldn’t be able to perceive at least as well as human drivers, if they have radar or some other type of sensor for their environment. Or even if they share info on the potholes they actually hit with the network.

    Legal concerns are huge, of course. If the number of people killed by cars could go from 40,000 a year to the mid-hundreds, it would seem we should be able to figure out a solution to the legal challenges, but it’s not hard to imagine those 300 or 500 deaths causing the whole transition to never be allowed.

  8. Submitted by David Markle on 11/14/2014 - 12:03 pm.

    A form of semi-automated transit

    Sounds kinda like sloppy PRT.

    • Submitted by Gerald Abrahamson on 11/14/2014 - 05:35 pm.

      It is the definition of PRT (personal rapid transit).

      PRT was always about getting the vehicle *close enough* to the individual AND to the desired destination to make it accessible and practical.

      The self-driving car literally comes to your house. Then it is able to deliver you as close to your destination as reasonably possible. Thus, the self-driving car is almost the ultimate in PRT–other than a transporter.

  9. Submitted by Rick Prescott on 11/14/2014 - 01:04 pm.

    The Alternatives

    Self-driving cars have many great hurdles to overcome — number one being the psychological disadvantage they face (people unwilling to believe they represent an advance).

    They also do little to reduce carbon emissions, and also have the potential to increase them if, as suggested in this post, people enjoy the convenience enough to begin living even farther from their workplaces than they do now.

    This will sound a little crazy, but the real innovation which has the potential to upend transportation patterns (and laws) is personal aircraft. These big brothers to the drones used by hobbyists have all the advantages of self-driving cars mentioned here without any of the liabilities (including many of those associated with weather).

    Already, air-busses are in development in Europe, as are two- and four-seat versions. The ownership of personal vehicles could virtually disappear as “air taxis” require only minimal landing areas, and no parking whatsoever. Even the infrastructure is largely digital. Automated air traffic systems could dispatch and route vehicles using technologies which are already mature.

    Like I said, this sounds a little crazy right now, but it actually has far greater potential than self-driving cars, which will always face backlash from passengers who will feel like their freedoms are being taken away.

    • Submitted by Bill Lindeke on 11/14/2014 - 01:54 pm.

      Energy costs make that implausable

      IMO, the amount of energy needed to fly makes personal aircraft on an everyday basis far out of reach for anyone but an elite class. Just not efficient way to move around.

      That said, it’s long been a dream. Frank Lloyd Wright’s broadacre city was predicated on personal helicopters for everyone.

  10. Submitted by Todd Piltingsrud on 11/15/2014 - 07:13 pm.

    autonomous cars don’t solve enough problems

    Consider the problems we have with cars now: they’re dangerous, they use too much energy to move too few people around, they’re expensive, they necessitate paving over huge tracts of some very expensive real estate, etc.

    Robot cars would be safer, no doubt. But they’d still use a lot of energy compared to other alternatives. (And making them electric doesn’t fix that.) Buses get much lower gas mileage, but they more than compensate by moving more people. Same with trains. Takes a lot of electricity to move a light rail, but they move a LOT more people. Cars can’t even come close to replicating this kind of efficiency.

    And then there’s the problem of highway infrastructure. If we don’t reduce our dependence on cars, we’ll still be stuck paving over large tracts of real estate.

    And lastly there’s the expense. It’s safe to assume that a robot car would be more expensive than a manual one. People assume we’ll solve this problem by sharing them. But wouldn’t a shared robot car still be more expensive than a shared manual car? If people were given the choice between two car sharing programs: one with robot cars and one with manual ones, which would most people choose? And before you answer, consider that for most people transportation costs are second only to housing costs.

  11. Submitted by Paul Udstrand on 11/16/2014 - 06:25 pm.

    Could smooth traffic out but…

    A lot of our traffic jams are the product of human psychology. For instance there’s absolutely no good reason for the slow downs on 494 by Penn Ave and Best Buy Headquarters. There’s just something about that little dip and jog in the road that makes drivers with human brains slow down. You see the same thing on 394, and 169, and the stretch of 94 between Riverside Ave and 280. Robots wouldn’t react psychologically to little dips and twists like that, so in theory, those jams wouldn’t disappear.

    But, the transition period would be a killer. For a decade or more you’d have robots on the same roads with human drivers and crashes would continue to happen with both kinds of cars. The whole outcome could be decided by that transition period. If people perceive the robots as a step forward, they might fly. If on the other hand people react negatively, and that could happen for a variety of reasons, robots would not take over. I mean look, the natural assumption would be to equate robot cars with an airplanes autopilot, but pilots don’t let the autopilot take-off or land the plane do they? One thing it is to let a car parallel park itself, but letting it drive you through tangletown is a different story.

    The other thing is cost of the cars themselves, I doubt these things would be cheap and unless household incomes start climbing it’s unlikely such technology could be purchased by most people. I see a future where there may be robot cars, but they are much fewer in numbers, still share the roads with a few human drivers, while most people use transit of some kind.

Leave a Reply