Watch Uber’s Self-Driving Trucks Moving Freight Across Arizona

The autonomous truck is already here. On Tuesday, Uber revealed that its vehicles are already moving freight across Arizona, the first steps toward a model where self-driving vehicles are used to deliver loads along highways with human drivers taking over at either end.

“We’ve been really hard at work for the last several months improving the technology, improving the performance of the system, and also working on building a product that can exist in the real world and solve real problems,” Alden Woodrow, Uber’s self-driving truck product lead, told reporters. “The team is hopeful that its solution will “make truck drivers’ lives and jobs better.”

The truck uses the technology under development by the Uber Advanced Technology Group, the team also working on the company’s autonomous taxi project. In June 2017, the company announced it was dropping all Otto branding, the autonomous truck company it acquired in August 2016. It also planned to use a 64-channel spinning lidar array on Volvo vehicles using Uber’s in-house software stack. The Uber trucks can use the mapping data accumulated from the cars, meaning time spend on the road benefits the vehicles across the board.

In the above video, the load starts off in the midwest and is driven by a human driver to a transfer hub in Sanders, Arizona. It’s then transferred to an autonomous truck that has a human driver ready to take over just in case, with the first truck taking a load back from the transfer hub. When the autonomous truck reaches the second transfer hub in Topock, Arizona, it’s passed back to a conventional driver, who takes it to its final destination in Southern California.

It’s a sample trip, but it’s indicative of the hub model Uber has in mind with these journeys. The end goal is to have the autonomous trucks running round the clock, moving loads in both directions between the transfer hubs. The system uses the Uber Freight tool that matches trucking companies with shippers, so the human drivers at either end are matched up via this tool. The person sitting behind the wheel of the autonomous truck is an Uber ATG employee, with the company seeking out truck drivers with years of experience to monitor the movements.

Watch an animated explainer below.

Right now the vehicles are only operating in Arizona and handing off close to the state border, the reason largely being that it aligns with the geography under testing right now. The team has faced a number of challenges testing on the highway, with unexpected construction and accidents just some of the issues the developers must consider.

“Truck drivers are very concerned about how other vehicles on the road drive around them,” Woodrow said. “It’s actually very stressful to be around vehicles that are driving unsafely, maybe unintentionally, but driving unsafely. Our self-driving trucks face the same challenges.”

There’s still some unanswered questions about Uber’s project. Right now, the company is not releasing specifics on how many trucks it’s using or miles it has completed, instead describing it as a “small fleet.” It’s also not releasing data about how often the human drivers are taking over from the autonomous system. Uber previously came under scrutiny after a leaked disengagement report in March 2017 seemed to show a higher takeover rate than Waymo, a comparison the company disputes as different companies use different reporting methods.

But as for whether you’ll one day see Uber-operated trucks whizzing around the country, right now it’s not the company’s goal. The trucks aren’t being designed to reverse into docks or anything else, with a focus instead on the transfer hub model. In the future, the team sees Uber’s role in the industry as one focused on partnerships.

“There are a lot of companies in the value chain that are very good at what they do, and so in general our desire is to partner with companies in the industry and find the right way for us to enable self-driving technologies,” Woodrow said.

It could come sooner than you think.