For the developers of self-driving vehicles, semi-trucks are a potential low-hanging fruit. Although big rigs are imposingly, intimidatingly huge, they also predominantly run on freeways and other fixed routes that are simpler to automate. Fleet managers are easier to convince with rational, financial arguments than individual car buyers. Plus, there’s a shortage of some 50,000 truck drivers in the US.

Today, a new startup, Kodiak Robotics, is edging out of stealth, announcing $40 million in financing, and telling the world it’s going to pick and ship that fruit. The company is worth watching because it’s cofounded by Don Burnette, who also cofounded Otto, a trucking startup acquired by Uber in 2016, and particularly notable because it led to a high-profile legal spat in the nascent autonomous vehicle industry. Uber eventually settled, and then shuttered its autonomous truck program last week. (Burnette had already left the company back in April.)

The autonomous world is a small and incestuous one, and fond of rebranding, so it can be tough to keep up. Here’s the Cliffs Notes. Burnette was originally at Google (now part of Alphabet), working on its self-driving car program (now Waymo). He left to cofound Otto in 2016, along with fellow Googlers Anthony Levandowski and Lior Ron. After Uber bought Otto for $680 million in stock, Waymo sued, alleging the theft of trade secrets. Uber paid out $245 million in stock to settle, but has now announced that it wants to focus solely on autonomous cars, not trucks. Its testing efforts will will now be under increased scrutiny after a fatal collision in Tempe, Arizona, in March, when one of its self-driving cars hit a pedestrian.

That leaves one fewer competitors in the self-driving big-rig market, and room for a new round of startups to flourish. But Kodiak isn’t just a rebooted Otto.

“I think there are a lot of aspects of the industry that are different from the days of Otto,” says Burnette. Chiefly, sensor technology has greatly advanced in the last 2.5 years. Back then there was only one company dominating lidar: Velodyne. (Lidar is a laser sensor most experts believe is crucial for self-driving cars to perceive the world. Indeed, it was designs for a new lidar sensor that formed the cornerstone of Waymo’s complaint against Uber.) Now, there are tens of companies building lidar, as well as cameras, sonar, and radar sensors, for autonomous vehicles.

Burnette also says he and the new team have learned a lot more about the industry they’re targeting. Instead of just throwing autonomous car engineers at the problem, they’ve spoken to truckers and shippers to understand the logistics business and the challenges. “It’s a much more mature view,” he says.

Kodiak isn’t ready to talk in detail about its business plan yet, but Otto’s model was to build affordable aftermarket systems to bolt onto existing trucks. Its promise was a $30,000 price tag.

Kodiak’s approach is likely to be similar, and the company is talking to sensor-makers to buy in what it needs. “We take a holistic approach; we want to use all the different sensors that are available,” Burnette says. “We’re not trying to build this super cheap or make it super expensive. We’re more trying to take advantage of all the modalities that are available to us today and combine them in an intelligent way.” The company’s unique selling point will be the integration of all that sensing, and the computing necessary to make sense of it, and decisions on driving.

Kodiak will be joining other startups like Starsky Robotics (which performed a fully-driverless truck test in Florida in March), Peloton (which wants to platoon trucks like giant land-trains), and Sweden’s Einride, which is so convinced it won’t need a human on board, its haulers don’t even have cabs.

The established truck industry isn’t waiting it out in a rest stop either. Volvo and Daimler are working on prototype autonomous versions of the rigs they already build and running tests in Europe.

Self driving car development feels like it's hit a plateau, as early promises pale against the hard reality of how to engineer a vehicle to deal with all the vagaries of the open road. But Kodiak says semi-truck development could move faster. For starters, freight won’t complain as much about a slightly jerky ride. And if engineers focus on highway and interstate driving, that constrains the problem by removing cyclists, pedestrians, and traffic lights. “You get on the road, stick in the right lane, and don’t hit the thing in front of you,” says Burnette.

That’s certainly a good start, but not nearly enough to replace the truck driver. Kodiak and others can build machines that drive from on-ramp to off-ramp, changing lanes themselves with minimal supervision. The human driver is still necessary. Except now those drivers can catch up on paperwork or sleep or even be located in a call-center to dial in to remote-control the tricky bits, helping logistics companies deal with the shortage of drivers. In short, they're still necessary if the shipping world wants to keep on truckin’.


More Great WIRED Stories

Read more: