Nissan and NASA join forces on self-driving robot car
Nissan’s end goal is to provide a complete system of seamless mobilityAutofile Staff
Published: February 1, 2018, 4:00 PM
Updated: February 5, 2018, 5:04 AM
By Peter Lyon
Walking into the backyard testing area of Nissan’s Research Center in California’s Silicon Valley, Dr. Liam Pedersen reveals his ‘baby’. “This Leaf may look like a car, but it is, in fact, a robot,” he says.
The former NASA robotics researcher, who was instrumental in creating the Mars rover as part of his work on future planetary exploration, is now Nissan’s principal researcher for autonomous vehicle development.
“What do our Nissan Leaf autonomous drive prototype and a Mars rover have in common?” he asks. Then answers, “They both use similar autonomous systems to navigate their respective environments. Robots on Mars, like the 4WD rover, need to sense their surroundings. They need to make their own decisions about what to do. And that is what a robot car also needs to do on Earth.”
Robot cars – perhaps more commonly called autonomous vehicles (AVs) – have intricate cameras, sensors and software and it is Pedersen’s team’s responsibility to create the Artificial Intelligence (AI software that goes into making the safe driverless cars of the future.
Joint work with NASA
Nissan reiterated its commitment to self-driving mobility at the recent CES 2018 by announcing that Nissan North America had agreed to continue its joint work with NASA Ames Research Center in California’s Silicon Valley.
The two collaborate on research and technology development for future autonomous mobility services, including a working demonstration in Silicon Valley. Nissan is the only automaker collaborating with NASA on next-generation self-driving technology.
Getting down to basics, a question many people may have is, why build autonomous driving vehicles at all? “Safety is our first issue,” says Pedersen.
“Each year, there are over 20,000 road fatalities in the U.S. and the victims are mostly under 25 years old. We can improve that. Road congestion is another concern. I spend two hours a day stuck in traffic. So a broader application of autonomous mobility would alleviate that congestion and shorten our commuting time. Autonomous vehicles also enable new transportation models. With shared transportation, EVs (Electric Vehicles) become more economical.”
In creating this new safer, more-efficient, cleaner world, Pedersen believes that electric vehicles and autonomous driving are mutually beneficial. “They are a set,” he says. “By employing self-driving technologies, we can make EVs more economical and emissions cleaner. Our aim is zero emissions and zero fatalities.”
Human interaction essential
Dutch director of NRC and former NASA engineer Maarten Sierhuis says, “It will be impossible to have autonomous vehicles driving around without ever needing help. Any autonomous system is built by humans for humans, so vehicles will have to interact with human beings.
“To do that flawlessly, Nissan has designed a system for seamless integration,” he explains. “It’s called SAM or Seamless Autonomous Mobility. It’s a support system for autonomous systems and transportation networks. We can liken it to air traffic controllers who manage thousands of airplanes in the air, and still, we have humans at a distance managing and observing the air space.
“Our system is very similar to this. SAM uses multiple cameras, LIDAR technology, laser range finder, Milli-wave radar and other onboard systems to enable it to seamlessly interact with society.”
Pedersen explains that a critical issue is how to be social in our society. “Just say you come to a four-way intersection. You have traffic lights, pedestrians crossing the road, cyclists, people in cars saying I got here first. A human can cope easily. But for a robot this is difficult,” he says.
The robot has to learn how to interact with people and other cars in a socially acceptable way. If it does not drive in an acceptable manner, people around it will start driving in unsafe ways to avoid it.
Acceptance is critical
“Robot cars have to be accepted by the community,” explains Pedersen. “The most important thing is that we do not cause inconvenience to the people around us. “Being too polite, too slow, too hesitant is unacceptable and unsocial. We must avoid this.
“Take this example,” he suggests. “At a road construction site, you may have a worker waving you to drive through. But at same time you have a red traffic light saying stop. What is a robot supposed to do in that situation?”
Drivers come across many situations like that each day, and the most important thing from Nissan’s perspective is that the robot car always has a way of getting out of this kind of situation. For every possible situation that may arise, the car must always have the option of calling a remote human for help.
Like a kind of Big Brother, the human-operated monitoring system always knows where the cars are located and whether they may need help. So the car can, in effect, call a remote location and get feedback on how it should proceed to clear the sticky situation it is in.
What happens, when an autonomous vehicle using these systems detects an obstacle while driving, is that data is transferred to a human mobility manager in a remote location who then paints a new path around the obstacle.
Once this new path is set, AI will then transfer that information to other vehicles enabling them to autonomously follow the same path without further human intervention.
According to Pedersen, a similar system is being tested by Nissan in Japan, but on local expressways where there are no traffic lights, no pedestrians and all cars are going in same direction.
“In Silicon Valley ,however, our job is to cater to the much more complicated local street situation where you have traffic lights, pedestrians, cyclists and cars going in all directions,” he says.
Seamless mobility the goal
Nissan’s end goal is to provide a complete system of seamless mobility. Pedersen says that eventually the system will manage thousands of vehicles at once creating a larger ecosystem for autonomous mobility. These self-drive technologies will drive services like robo-taxi, robo-delivery, robo-shuttle and robo-sharing.
There is no denying it anymore. Autonomous vehicles are coming, and sooner than we may think. This R&D partnership with NASA allows Nissan’s intelligent mobility to safely overcome the multitude of technical hurdles as it prepares for a world where self-driving vehicles will transform society for all people.
It all sounds like something out of a James Cameron movie. Hugely beneficial, but a little scary all at once. And while I know thousands of people will love to take handy self-driving taxis in the near future, I just wonder about self-driving pizza deliveries.
How will they bring the pizza to your door if there’s no human to physically carry it? Perhaps that will be the next big challenge for robot engineers.