A Driverless Car Race From Rome To Shanghai
ROBERT SIEGEL, host:
A grueling road test of a different kind begins next week in Europe. It's an 8,000-mile auto trip from Rome to Shanghai.
These aren't just any cars. They are two, solar-powered, electric vehicles operating autonomously - that is, they have no drivers. Instead, the cars, which are really vans, are tricked out with laser scanners, cameras, GPS sensors and lots of computing power.
And joining us to explain the how and the why of this unusual road trip is Alberto Broggi, professor of computer engineering at the University of Parma. He directs the school's VisLab, which has been developing driverless vehicles for more than 15 years. Welcome to the program, Professor Broggi.
Mr. ALBERTO BROGGI (Professor of Computer Engineering, University of Parma): Hello, thank you very much.
SIEGEL: How does it do it, in a word? That is, how does the van read enough of its environment to be able to stay in its lane and avoid any obstacles, stationary or moving?
Mr. BROGGI: So we have cameras on the vehicle. We have seven cameras. And we have laser scanners. By fusing the information coming from the cameras and laser scanners, we try to reconstruct the 3-D world which is outside.
SIEGEL: But you will not only drive across mountain ranges and stretches of desert, youll also be driving in cities, for example getting out of Rome. That sounds like quite a challenge for a driverless car.
Mr. BROGGI: Yes, actually, we have a small trick to do that. What we do is that we have two vehicles. So the first one will be manned, let's say partial autonomous and partial manual. And this vehicle will be driven through cities, towns and then a long stretch of road in Siberia and whatever, and the second vehicle will be following it.
So the second vehicle will be the really autonomous vehicle. And if the first vehicle is visible, then the second vehicle will follow it. If the first vehicle is not visible because it's maybe too far away or there is another car in between, then the first vehicle will be broadcasting its GPS position, so that that second vehicle will be able to roughly follow the trajectory planned by the first vehicle.
SIEGEL: How close are you to having this down? That is, would you trust this ride? Would you put your children in the seatbelts in the backseat and trust the autonomous vehicle to know when there are obstacles, when there are red lights, when there are pedestrians jaywalking?
Mr. BROGGI: You know, the greatest challenge is always perception, to understand what we have in front. And perception is really, really difficult.
Think about a pedestrian detection system. The pedestrians have different shapes, different colors, different postures. So thinking about any variance of a pedestrian is very, very difficult.
SIEGEL: And a human driver seeing, say, a soccer ball roll into the road might assume there could be a child following the ball into the road, but that's not something an autonomous driver might be able to calculate.
Mr. BROGGI: Yeah, you're right, you're right. And there are also other functions that a human makes usually and the robot is not able to do right now. So we are perceiving what we see, but we're not perceiving what we cannot see, and what we need to infer.
SIEGEL: Professor Broggi, what is a possible application of this technology beyond the idea of having driverless cars someday that you might imagine in the reasonably near future?
Mr. BROGGI: The first that I always mention is in the agricultural domain. So think about having your tractor on your field moving autonomously for 24 hours a day, seven days a week.
There are companies which are interested in putting, you know, this technology in other kind of vehicles, like road construction or mining or whatever. So any kind of vehicle is the target of this kind of application.
SIEGEL: Well, Professor Broggi, have a good trip from Rome to Shanghai or however far you might be along with it, and thank you very much for talking with us about it.
Mr. BROGGI: Thank you very much to you. Thank you.
SIEGEL: Professor Alberto Broggi is the project leader for the VisLab Intercontinental Autonomous Challenge. He's based at the University of Parma in Italy, and his team is going to test two driverless cars on a road trip from Rome to Shanghai.
NPR transcripts are created on a rush deadline by Verb8tm, Inc., an NPR contractor, and produced using a proprietary transcription process developed with NPR. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.