When you step on Olli, a bubble-shaped driverless bus now in trials in Knoxville, Tennessee, you can ask the bus if it will take you where you need to go or to recommend a restaurant for dinner, and the robot will politely answer. Depending on location, the electric minibus can also work like Uber, letting passengers request a ride through an app.
Olli, which is part of a series of pilots and demos now, hopes to run commercially in Las Vegas in 2017, making it one of the first in the U.S.
In part, self-driving buses are coming to the market faster than cars because they are attempting only to solve the "last mile" problem—ferrying people short distances between a train station and home, for example—and they don't ever drive faster than 25 miles per hour.
Plastered with 30 sensors, from lidar and radar to cameras, the slow-driving Olli can spot pedestrians far earlier than a human driver, and come to a full stop 10 times faster.
"If there's something in the way, something it's unsure of, if there's any kind of technical problem, the vehicle stops and a human remote supervisor will be notified and human intervention can then come and take over," says Corey Clothier, VP of product management at Local Motors, the company manufacturing the bus. "That's very different than being on the highway going high speeds in the middle of nowhere."
Some cities and organizations, such as airports, plan to use the minibus on a fixed route, like a traditional shuttle. Others hope to use the on-demand app to let passengers request a pickup, making public transit more convenient.
Because the bus is electric, it has a smaller carbon footprint than most buses on roads now; because it's driven by a robot and not a human, it's also an extra 10% to 25% more efficient. Next year, the company plans to begin testing solar panels on the roof of the shuttle, so it can partially charge itself as it drives.
Many of the parts on the bus are also 3D-printed, so they can be recycled when the bus is replaced. "We wash it, shred it, and it's ready to be integrated in the raw material for printing again," says Clothier.
Since the routes will be in a semi-constrained, slow environment where the bus can easily stop if it needs to, the company says it's fully safe. The bigger obstacle to getting it in use in more cities is acceptance by humans.
"The biggest challenge is people," he says. "The technology's ready to go. It's been ready—at least in these type of environments—for a little while, for the last couple of years."
Humans may be slower to adapt. At a recent test on roads in Detroit, drivers were annoyed that the bus was going the speed limit. "People were mad at it, throwing apples at it," he says (Update: Clothier now says this didn't happen, but was a hypothetical). At other demos, Clothier says people have "bullied" the bus, jumping in front of it to test the brakes, and creating a jerky ride for everyone inside.
"We don't always treat machines properly, even if humans are on board," he says. Through demos—and a series of pilots in the cities that plan to use the bus—the company plans to sort out the human interaction challenges, both for other drivers and for passengers. And then it expects that, along with other companies, the self-driving bus as transportation will quickly expand.
"I think there will be a small handful [of commercial deployments] next year, and in 2018, it will double, triple, expand five times," he says. "Then I think it will go from there."
Have something to say about this article? You can email us and let us know. If it's interesting and thoughtful, we may publish your response.