Why Autonomous Trucks Are Harder Than Everyone Expected

One company leading the charge these days is $Aurora Innovation (AUR.US)$, which has driverless test trucks on the road right now. It recently raised nearly half a billion dollars as it prepares a commercial launch of its technology by the end of the year. Yet getting to that point has hardly been easy for Aurora, despite being founded and staffed by veterans of Google's Waymo, Uber, Tesla and others.

While Tesla is racking up headlines this week as CEO Elon Musk aims to elaborate on why he's betting the farm on autonomy, it's hardly the only player trying to "solve" self-driving vehicles. And on an appearance of today's Pivot podcast with journalist Kara Swisher and professor and venture capitalist Scott Galloway, Aurora co-founder Chris Urmson elaborated on the challenges facing this space in particular.

Urmson admits that when a driverless car is operating in a city - where General Motors' Cruise and even the occasional Waymo robotaxi have seen high-profile mishaps over the past few years - there is "more to interact with." Construction, pedestrians, cyclists, other cars and so on.

Urmson added that many companies in the automated trucking space - he doesn't name them but they include Embark, TuSimple and Waymo—have either left that field or moved out of the U.S. Some competitors "didn't really understand the strategic investment you'd have to make," he said. For Aurora, that included its LIDAR system, which Urmson said "allows us to see much further than you can see or any of the, we think, the Robotaxi folks can see."

Urmson brought up one challenge the entire autonomous sector is dealing with: regulations. Right now, the laws around driverless cars, robotaxis, test trucks and so on are a state-by-state patchwork. Technically, he said Aurora can operate in 44 U.S. states, but since this is an interstate commerce issue he'd like to see a proper federal standard for the tech - an ongoing problem for everyone in the space.

This lack of regulation is also part of why robotaxi services like Waymo and Cruise only operate in certain places, or why Mercedes-Benz's hands-off, eyes-off Level 3 automated system can only be used in California and Nevada under certain conditions. As for Tesla's Full Self-Driving tech, it is the subject of a number of regulatory probes, lawsuits and even a federal criminal investigation. That system depends on cameras and AI, not LIDAR, but Urmson's co-founder Anderson recently discussed the difference he sees between the two approaches.

That's a big deal because, as this podcast points out, everything you see in the room around you right now was probably hauled on a truck at some point. The U.S. trucking industry moved USD987 billion worth of gross freight revenues just last year. And while Aurora's approach to automation may sound like bad news for those employed in the trucking space - a demanding but decent-paying path-to-the-middle-class job that doesn't require a college degree - the industry has been facing a driver shortage for years. Automating that sector could be a pathway not just for self-driving car tech, but also for keeping America's insatiable appetite for stuff running.