99% Invisible - 171- Johnnycab (Automation Paradox, Pt. 2)

Episode Date: July 1, 2015

More than 90% of all automobile accidents are all attributable to human error, for some car industry people, a fully-automated car is a kind of holy grail. However, as automation makes our lives easier and safer, it also creates more … Continue reading →

This is 99% Invisible. I'm Roman Mars. In 1956, General Motors hosted a car expo called Motorama. As with all car expos, Motorama was a chance to show off concept cars and other kinds of long shot projects that they hoped would revolutionize the auto industry. One of the most forward-looking things on display was in a car, but a movie. That's forward-looking producer Sam Greenspan. The film was called Key to the Future. In it, we see a family of four cruising along a desert highway in a beautiful futuristic GM Firebird.

It's got aerodynamic fins, gas turbine engines, and all glass passenger cabin. It looks like a car from the Jetsons. Only it's not flying. It's driving on a highway. But this car is actually just the set dressing for what GM really wants to show us. Their vision for a self-driving vehicle, you're now under automatic control. Hands off steering.

After being put on automatic control, the dad who's driving the car pushes the steering wheel out of the way. Oh, this is the life. Say, cool, comfortable. Mind if I spoke a cigar? No longer tethered to the steering wheel, Dad uses both hands to light his cigar. The car zooms along unimpeded. For nearly as long as there's been an auto industry, there have been dreams of a car that

drives on its own. A car where people can enjoy all the freedom that cars offer, but without the anoints or the dangers of driving oneself. I can't speak to what specifically inspired GM to dream up the self-driving car we see in this film, but it's very possible that it had to do with safety. The film is only nine minutes long, but they can't stop commenting on how darn safe this automatic driving is. The safe, easy way to make a turn. The year that this film was shown at Motorola, 1956, that year alone, there were nearly 38,000 vehicle-related deaths in the United States. Ever since the 1950s, there has never been a year

with fewer than 30,000 people killed in car accidents. Vehicle-related accidents are the number one cause of death in the United States for people between the ages of 5 and 34. And more than 90% of all automobile accidents are attributable to one thing, human error. So for some industry people, a fully automated car is a kind of holy grail. The thing about automation though, is that as it solves problems, it inevitably creates new ones. As automation makes our lives easier and safer, it also creates more complex systems and fewer humans who understand those systems, which means, when problems do arise, people

can be left ill-equipped to deal with them. Human factors engineers call this the automation paradox. Last week in our story about automation and aviation, we heard about various ways people are trying to deal with the paradox in that industry. For one, pilots are being encouraged to practice manually flying their planes to keep their skills polished. And engineers are also trying to make smarter, more collaborative automation that doesn't strip skills from pilots, but instead works with them to complete tasks. The position I would adopt is one that has been termed a human-centered automation, where we see what you need to do and make the automation a team player.

That's Nadine Sartre, a human-factor's engineer who believes human pilots still have a crucial role to play in flight. I don't think the answer to it is to say, let's just get rid of the human and let's fully automate the vehicle. You know, I do feel comforted by the fact it's a pilot up there. And that is Chris Irmson. I'm Chris Irmson. I lead the self-driving car project here at Google. Chris may be glad that there are pilots and planes, but he's got a very different vision for cars.

He believes the way to solve the automation paradox for cars is to take humans out of the operation entirely. If you have a steering wheel there, there's kind of an implicit expectation that you're going to do something with it. Where Nadine Sartre thinks that allowing for human intervention is best. Yes, you still think need a driver who is in the vehicle and who has the opportunity at least to intervene when necessary. Chris Irmson's goal is to make a car that is safe because no human is driving it.

You get to sidestep all of these kind of control confusion, kind of potential challenges by taking that out of the way. You get to remove a part of the system that you can't design. No human drivers, just human passengers. We are about to enter uncharted territory for automation. Automation has been appearing in cars almost since they hit the market. Antilock brakes, automatic transmission, power steering, cruise control, GPS, cars that can parallel park themselves,

the fully autonomous self-driving car has been a long time coming. There have been a few players developing self-driving car technologies, the most visible among them being Google. In 2009, Google began retrofitting Toyota and Lexus cars with new technologies, allowing the car to drive on its own.

The cars can accelerate, stop at traffic signals, make turns, merge onto freeways, and avoid pedestrian cyclists and other cars, all with little or no intervention from anyone inside the car. And then in 2014, Google started manufacturing cars of their own design, cute little two-seaters with no gas pedal, brake, or steering wheel. They're designed for the user to input a destination and just sit back and let the car do everything else. Google has videos that show blind people and children

sitting in what we normally think of as the driver's seat as the car scoots around a test track. Well, the intent is that you shouldn't have to have any training. You should be able to get in them, tell them where you want to go, tell them you change your mind and you want to go somewhere else and then get out. Google wouldn't show me the cars, but Chris did show them to planet money's Steve head. So you want to open it up and show me the inside?

Sure. So this is one of our early prototypes. Well, actually, it's our fifth generation prototype. This vehicle is one of the ones that will start to be testing on the roads in Mountain View in the next, over the summer. We talked to Steve about what it was like afterwards. When you get inside, there's no dashboard,

there's no steering column. So I'm about six feet tall, and the seat is all the way against the back window of the car, and I can stick my legs straight out in front of me because there's nothing there. Well almost nothing. What you can see in the center here, so this is a steering wheel breaking gas pedals and these are for our test drivers to use, but they're removable. California law allows autonomous cars on the roadways, but only if they can be driven by a human. But the design of this car is to have none of that, right? And so there are very, very few user controls inside.

There are buttons to lower the windows. There's a button to pull over to the side of the road and stop. You know, if you suddenly see a yard sale, you want to check out or a friend you want to pull over and talk to. But there's also an emergency stop button, a button that halts the car in its tracks. Kind of like on an elevator, you know how there's the big red button? Well, I really need to get out now, there's something very wrong here.

I've never pressed the red button in elevator, but it's kind of comforting to know that's there. The cars look kind of endearing and goofy, they're small and round. Steve refers to them as bubble cars, not that they would let him ride in one. Yeah, so they didn't let me ride around in the bubble car, and in fact the bubble car hadn't hit the streets of Mountain View, but he was able to get inside one of their older models. It was a retrofitted Lexus SUV.

Okay, should I jump in the front or the back? Probably. Yeah, probably this back left corner right here. I think the most remarkable thing for me was how quickly the experience became just kind of boring. It came to a full and complete stop at the stop sign. Definitely. So the car follows the law very closely. To Steve as a passenger, the car's driving was more or less indistinguishable from a human driver. Well, a human driver that always follows the rules of the road.

We were driving through a residential neighborhood in Mountain View, including through streets that were under construction. So it was navigating streetscapes that were changing. You know, lanes were blocked off, lanes were closed. And at one point on their drive, a construction worker was directing traffic, and Steve was chatting with the other person in the car. We were both sort of distracted. And suddenly, all on its own.

Temporary stop sign. Wow. The car slammed on the brakes. It was startling. So that was startling. So, that was quite interesting. Steve and the Google employees hadn't seen it, but the construction worker had flipped his handheld sign

from slow to stop. The machine had noticed that handheld stop sign when all of us had missed it, and it actually kind of scared all of us because we didn't see it. Even the Google employees hadn't seen that before. The autonomous car was working better than they had expected. Maybe these things can react faster than humans, but sensors will fail, navigation will

get screwy, onboard computers will break. Car owners barely know how to fix their own vehicles as it is now, so will people be able to fix the kinds of complicated problems that come with a self-driving car? It really depends on how they end up being used in the market, right? Chris imagines that if a self-driving car breaks while it's on the road, it'll pull over. And another car will come fetch the passengers, leaving the broken one for technicians to fix. Which means, Google doesn't want to just change how we relate to our cars.

Google wants to change how cars relate to society. If you look at the car that you own today, it's kind of sat parked somewhere 95% of the time, and that's kind of poorly using both your financial resources that went into it, but also the environmental resources, the material and energy we put into making that vehicle. So if we can use that more efficiently by sharing it, that sounds fantastic. Maybe, eventually, instead of owning these cars, you would hail them from off the street. And when you get out, the car would drive off and pick up someone else. In other words, it's possible that these cars could enable a world in which no one owns a car,

and all of us just get around by robot taxi. Well, so the great thing about predicting the future is that you're always strong. My name is Costa Samaris. I'm an assistant professor in the Civil and Environmental Engineering Department at Carnegie Mellon University. Costa Samaris has thought a lot about how self-driving cars will change society. He is the co-author of a report called Autonomous Vehicle Technology, a guide for policymakers. So Costa, like Chris Hermson, and pretty much everyone else I met working on autonomous vehicles, he thinks that these things hold a lot of promise.

So reducing crashes is the first way that automation will help society. But that's really only the starting point. After all, the vast majority of American cities are built around cars. And so, transforming the car means transforming the city, starting with parking lots. Because if all or most cars are taxis, we don't need to clear out spaces and congested downtowns just so cars can have a place to sit idle. Get out of the robot taxi and it's off to give a ride to

someone else. Urban cores get denser, downtowns get more walkable, and the roads that we do have can be used more efficiently. All the things we learned in drivers at, staying a certain amount of seconds behind a speeding car, those things will now be taken over by self-driving vehicles and they will make the determination of how far it's to be from the car in front of us and in order for everybody to be safe. This would allow cars to get closer together and not just bumper to bumper, side to side. The Department of Transportation requires that highway lanes be at least 12 feet wide. That's about twice as wide as the average car.

So suddenly, a three-lane highway can accommodate six lanes without any new construction. Less space between cars means more cars can get where they're going faster. But these cars could also reorganize cities for the worse. Right now I have a cost of driving, and that cost is running my vehicle, and also the time that I'm in it. And if one could read or sleep or write emails in a robot car, people might feel just fine about living far outside the city

and commuting a long way in. And that type of scenario, we would see lots more vehicle miles traveled, lots more sprawl into the deep excerpts and into the country. And of course, the inside of the car could become another place for advertising. Your route data could be for sale.

Maybe it knows I hadn't had coffee because it's connected to my refrigerator and my coffee maker at home and it says, he hasn't had coffee yet, let's make the slow route next to the coffee place. Or the route data could even become a matter of national security. Bad actors can take control of your car or a fleet of cars and stop every car that's on the bay bridge at one time for malicious intent. Clearly there are still a lot of details to work out. But I wonder, once the science is done and the policy is carefully considered, will people

even want these things? Is America ready to give up the steering wheel? I mean, will people still go on road trips and tailgate in the stadium parking lot? Will people even want these things? Is America ready to give up the steering wheel? I mean, will people still go on road trips and tailgate in the stadium parking lot? Will you still be able to get in your car and go on an aimless contemplative drive? I guess we haven't really thought about that.

That's Chris Irmson from Google again. My assumption is you would give it a destination and say where you wanna go, but you'll always be able to change it, right? So I don't do a whole lot of contemplative driving, but when I do, you know, you start heading somewhere and then you're like, oh, maybe I'll go this way afterwards. But you know, sometimes you really just need to get in your car and go. Like say, if you're getting chased by gun-toting corporate goons trying to stop you from aiding

Martian colonist rebels, you know, like Arnold Schwarzenegger did, in total recall. Hello! I'm Johnny Cat. Where can I take you tonight? violinist rebels, you know, like Arnold Schwarzenegger did, in total recall. At which point Arnold rips the robot driver out of the car and pilots it himself. It's funny because as fantastical as the sci-fi world is, we can recognize the same kinds of user annoyances we have here in the present. And we can see Arnold as heroic because he can do the things the machines can't. And think about Star Wars, when Luke Skywalker is on his way to blow up the Death Star. For one, his robot co-pilot gets fried.

And he turns off his onboard automation. Because automation is not as good as what Luke can do on his own. As much as we love building things that make our lives easier, it seems like we never get tired of seeing someone cast the robots aside. We love seeing people do things by hand. It's like we all have this anxiety about if we lose the ability to do something ourselves, are we losing a piece of ourselves or something? We are going to have to answer these existential questions about our cars kind of soon.

Chris Erbsen has said that his personal goal is to get the Google self-driving car done by 2020. So that his two sons, the oldest of whom is 11, will never need to get a driver's license. I would definitely think that when his son reaches 16 or 17, you would have to get a driver's license. Raj Raj Kumar is friends with Chris. They used to work together at Carnegie Mellon University before Chris went over to Google. Raj is co-director of CMU's Autonomous Driving Collaborative Research Lab. The university has been working on automated cars since the 1980s. About six months back, we literally celebrated the 30th birthday of actually working on

self-driving vehicles. Raj is also the CEO of a company called Automatica, which has a prototype vehicle that has already driven itself from San Francisco to New York, but Raj doesn't plan on taking the steering wheel out just yet. In fact, his car actually monitors the person in the driver's seat and beeps at him or her if the driver's eyes wander from the road. Raj is not in a rush to have us 100% dependent on his autonomous technology. The way he sees it unfolding is with incremental advances over time.

Just like right now, pretty much all cars have cruise control and some cars have adaptive cruise control, which calibrates speed. The next generation of this technology, I guess, would be you do not have to control the steering wheel either. You get onto the highway, pick a lane, and then engage this so super I guess, would be you do not have to control the steering wheel either. You get onto the highway, pick a lane and then engage this so super cruise control and then you'll likely steer itself, break itself and accelerate itself. Raj also imagines an automatic mode from when you're stuck in a traffic jam. And then the number of scenarios that are automatable will increase over time and one fine day,

the vehicle is able to control itself completely But that last step will be a relatively minor incremental step and one would barely notice that hey, this actually happened So to me it will take anywhere between 10 to 20 years So while Raj does see humans as having an important role to play as we transition toward fully automated vehicles His goal like Chris Irmsons, is to eventually get human drivers out of the process entirely. But even when that day finally comes, when we don't need steering wheels and don't even notice that they're gone, there will still be accidents.

There will always be some edge cases where things do go wrong, beyond anybody's control. This is kind of the nightmare scenario for all of us researchers. If somebody deploys something prematurely, and then it turns out that it ends up killing a child, God forbid. If and when people do get hurt in autonomous cars, there may be circumstances in which the passengers wouldn't have been injured had there been a human at the wheel. These cases will be hard to reckon with, but we'll have to keep in mind that there were 30,000 Americans dying every year in human-driven cars.

If autonomous vehicles lead to fewer car accidents, then as with planes, we're going to need to accept edge cases and periodic failures as the cost of living in this safer world. But I don't wanna be an edge case. I don't want to be an edge case. I don't want to give up total control of something when being in control might save my life. And yet while I was reporting the story, I almost caused a terrible car accident. It was early, the sun was in my eyes, and I wasn't really paying attention as I had made

a left through a crosswalk. And I nearly struck a pedestrian. I got so close that the guy could have reached out and slammed his fist on the hood of my car, which he didn't do, though I wouldn't have blamed him if he had. And this terrifies me. I want automation to save me and everyone else from myself. And I want automation to protect me from everyone else. And so when the time finally comes, I will probably roll the dice on autonomous cars. I think we all will eventually.

Maybe one day I will be an old man bouncing grandkids on my knees saying, you know, when I was a young man, we drove cars ourselves. You move them around with foot pedals and a thing called a steering wheel. And they'll say, well a steering wheel. And they'll say, well, that sounds amazing! And I'll say, eh, it was alright. 99% Invisible Was Produced This Week by Sam Greenspan with Katie Mingle, Avery Troubleman and me Roman Mars, reporting from Sweden. Extra special thanks to Steve Han and Planet Money, the great, great Planet Money,

Tristan Cook, Bob Sharet, Chris Nostle, Joe Brown, Caitlin Jabari, Jackie Miller, and Sherry Stokes. This story started life as another story about automatic doors in trains in Argentina, special thanks to the folks who helped us out there, Tomas Balmulseta, Gisermo Fiorita Catalano, Mariano Pachala, Annie Murphy, Jim Greller, Anna Adlerstein, Stephanie Foo, Ernesto Falzone and Ricardo Barriero at the Association of Trainway Friends in Argentina.

We are a project of the 99.7KALW San Francisco and produced under the offices of Arxan, a beautiful architecture firm in beautiful downtown Oakland, California. You can find this show and like the show on Facebook, all of us are on Twitter, we're on Instagram, Tumblr, we even have a Spotify playlist, but you can find all the information you need about this show and every other episode we've ever done at 99pi.org. ever done at 99pi.org. Radio to P.O. from PRX.