Uber Is Betting We’ll See Driverless 18-Wheelers Before Taxis
In a battered warehouse in San Francisco, Uber is working on what it thinks will be a shortcut in the race to make money from vehicles that drive themselves. A fleet of six modified white Volvo truck cabs operate out of a brick building in the SoMa district popular with technology startups. Around the clock, at least one of the vehicles is steering itself around Bay Area highways.
The trucks have radar, cameras and lidar—which maps in 3-D using lasers—added to their roofs and fenders by Otto, a startup that Uber acquired last month. The startup’s team is sharing data and technology with Uber’s group in Pittsburgh, which is working on autonomous cars to carry passengers. But Otto is still focused on its original business plan—creating a computer copilot that can let a trucker sleep during long stretches of highway driving. The truck would pull over and stop when it was time to leave the highway, or for the driver to take over again.
Otto’s founders claim that will make transporting goods safer and more efficient, creating a new source of revenue for Uber. And they argue that limiting their vision to trucks and highway driving means they can make real money out of autonomous vehicles much sooner than companies committed to passenger vehicles for robotic taxi service, such as Alphabet. Ford, Uber, and BMW have said they will be ready to operate robot taxis within five years, but experts predict the challenges of urban driving will limit them to small, controlled areas.
“It's still a very hard problem, but all the building blocks are there, and it’s much simpler than city driving,” says Lior Ron, who cofounded Otto and for five years led Google’s maps project. “We can show autonomy sooner rather than later, showing the path for the rest of society.”
Driving on highways removes the need to teach software to understand the unpredictable obstacles and social situations that arise on urban streets with crosswalks and four-way stops. It also simplifies the task of creating the detailed 3-D maps that autonomous vehicles rely on because fewer roads need mapping and highways are less complex than urban streets.
Ron says there is also less pressure to shrink the cost and size of sensors and computing equipment for a giant commercial truck than for a car. The company’s decision to have trucks pull over—perhaps in dedicated rest areas—before asking a human to take over avoids the risk of a person not being ready to do so, something experts say make partially automated systems like Tesla’s Autopilot a bad idea.
Ron cofounded Otto with Anthony Levandowski, an originator of Google’s self-driving car project, and two other ex-Google engineers in January. The company now has nearly 100 employees, including other veterans of Google’s autonomous driving project. Nearly 20 are safety drivers who sit behind the wheel of trucks during testing, ready to take over should Otto’s software go wrong.
Steven Shladover, a researcher at the University of California, Berkeley, who has worked on automated driving, says that while Otto’s strategy removes some challenges faced by passenger car projects, it adds new ones.
“The consequences of a software bug causing a 40-ton truck to veer out of control are far worse than the consequences of a similar bug on a two-ton car, which is likely to be terrifying for the general public and their elected representatives, who oversee the safety regulators,” he says.
Uber has earned a reputation for ignoring or antagonizing regulators of taxi cabs, but Ron says he will be working closely with state and federal road safety agencies for every step of Otto’s development plans. Several states, including Nevada, have already invited Otto start testing on their roads, he says.
Otto’s trucks have already driven themselves many tens of thousands of miles on public highways, and are also being tested with challenges such as obstacles and bumpy pavement at a proving ground in the Bay Area.
Ron says the company wants to add its kits to trucks owned by a trucking company to test and refine the technology on real routes by the end of the year. Partnerships like that, he says, will eventually provide the data to show that Otto’s technology is safer than the average truck driver and should be allowed to go to work.
Geoffrey Hinton tells us why he’s now scared of the tech he helped build
“I have suddenly switched my views on whether these things are going to be more intelligent than us.”
Meet the people who use Notion to plan their whole lives
The workplace tool’s appeal extends far beyond organizing work projects. Many users find it’s just as useful for managing their free time.
Learning to code isn’t enough
Historically, learn-to-code efforts have provided opportunities for the few, but new efforts are aiming to be inclusive.
Deep learning pioneer Geoffrey Hinton has quit Google
Hinton will be speaking at EmTech Digital on Wednesday.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.