Trucking is already being automated, but the full picture is more nuanced than a simple yes or no. The first fully autonomous freight corridor in the United States launched in March 2025, connecting distribution hubs in Texas and California. Aurora planned to begin commercial autonomous trucking service in Texas by April 2025. These are real trucks moving real freight on public highways without a human behind the wheel. But the technology has clear limits, and the shift will be gradual, partial, and shaped by weather, regulation, and the physical realities of getting cargo from a warehouse to a loading dock.
How Autonomous Trucking Actually Works
The model taking shape isn’t one where a driverless truck picks up a load in Miami and drops it off in Seattle. Instead, the industry is converging on what’s called an Autonomous Transfer Hub Network. Human drivers handle the first and last miles of a route, picking up freight from warehouses, navigating city streets, and backing into tight loading docks. At a transfer hub on the edge of a metro area, the trailer is handed off to an autonomous truck that drives the long highway stretch in between. At the destination hub, another human driver takes over for final delivery.
This middle-mile focus makes sense because highway driving is the most predictable part of trucking: limited access points, consistent lane markings, fewer pedestrians and cyclists, and mostly straight roads. It’s also the segment that puts the greatest strain on human drivers through long monotonous hours that contribute to fatigue-related crashes.
Where Testing Is Happening Now
Most autonomous trucking pilots are concentrated on high-volume freight routes in the southwestern United States, where weather is dry, roads are well-maintained, and lane markings are reliably visible. Texas has emerged as the primary testing ground, with multiple companies running routes there. Smaller-scale testing of truck platoons (where automated trucks follow a lead vehicle in close formation) has taken place in North Dakota, and autonomous box trucks have been tested in Arkansas and Toronto.
The geographic concentration isn’t random. It reflects a fundamental technical constraint: these systems perform best under clear skies on well-mapped highways. Companies that once pursued broader ambitions, including Ike, Otto, Peloton, Pronto, and Starsky Robotics, have all exited the autonomous trucking space, a reminder that commercial viability has been harder to achieve than early hype suggested.
The Weather Problem
Adverse weather is the single biggest technical barrier to widespread autonomous trucking. Fog, heavy rain, snow, and dust can severely degrade the sensors these trucks depend on. Lidar, the laser-based system that builds a 3D picture of the road, works extremely well in clear conditions but cannot function accurately once the road is covered in snow. Cameras, which detect lane markings and read signs, are most affected by fog, heavy rain, snow, and dust.
Radar can see through precipitation and detect large fast-moving objects, but it can’t operate independently to handle all the safety tasks a truck needs. Falling snow can trigger ultrasonic sensors to beep continuously with false obstacle warnings when nothing is actually there. One workaround is using pre-built 3D maps to help the truck orient itself when lane markings are invisible, relying on above-ground landmarks instead. But this is a patch, not a solution. Routes through the northern plains in January or mountain passes during a storm remain well beyond current capability.
Safety Compared to Human Drivers
The safety case for automated driving is strong in some respects and complicated in others. Only 1.8% of autonomous vehicle accidents are attributed to inattention or poor driving behavior, compared to 19.8% for human-driven vehicles. That’s a massive gap, and it reflects the obvious advantage of a system that never gets drowsy, distracted, or drunk. One widely cited estimate suggests that if autonomous vehicles were introduced with safety performance just 10% better than an average human driver, roughly 600,000 fatalities could be prevented in the U.S. over 35 years.
But autonomous systems have their own weak spots. They get into accidents at significantly higher rates in work zones and areas disrupted by previous incidents like disabled vehicles or spilled cargo. About 5% of autonomous vehicle accidents happen in these conditions, compared to 1.3% for human drivers. Autonomous vehicles also struggle more during dawn and dusk lighting, with accident rates 5.25 times higher than human drivers in those conditions, and during turns, where their accident rate is roughly double.
There’s an ironic pattern in rear-end collisions, too. In 79% of rear-end accidents involving an autonomous vehicle and a human-driven vehicle, the human rear-ends the autonomous vehicle. The automated truck is doing what it’s programmed to do, slowing or stopping precisely, and the human driver behind it isn’t expecting that behavior. This suggests a transition period where mixed traffic creates its own risks.
What Federal Regulators Are Doing
The regulatory framework is still catching up. The Federal Motor Carrier Safety Administration is considering changes to its rules to account for the differences between human operators and automated driving systems, but neither federal vehicle safety standards nor trucking-specific regulations currently include technical requirements for the autonomous technology itself.
For trucks with lower levels of automation (where a human must still be present), existing rules around commercial driver’s licenses, drug and alcohol testing, medical qualifications, hours of service limits, and bans on handheld devices all still apply. The bigger question is what happens with fully driverless trucks. FMCSA is exploring how to adapt its inspection and maintenance rules for these vehicles, acknowledging that current regulations weren’t written with them in mind. The agency’s approach so far has been to modify existing rules rather than create an entirely separate regulatory framework.
All commercial trucks, automated or not, must still pass a comprehensive inspection at least once every 12 months and be kept in safe operating condition at all times. That requirement isn’t going away, but the checklist will need to expand to cover sensors, software, and communication systems that don’t exist on traditional trucks.
What This Means for Trucking Jobs
The U.S. Department of Transportation has acknowledged that automating the driving task would decrease the number of driving jobs, particularly in long-haul routes. But the picture isn’t purely one of elimination. Several new job categories are expected to emerge. Remote managers would supervise autonomous trucks from control centers, monitoring multiple vehicles simultaneously and intervening when the system encounters something it can’t handle. Technicians would maintain increasingly sophisticated sensors and software. Data management and cybersecurity roles would support the digital infrastructure behind the fleet.
Some driving jobs would likely shift rather than disappear. The hub-to-hub model still requires human drivers for first-mile and last-mile work, including the complex maneuvering in urban areas, construction zones, and facility lots that autonomous systems handle poorly. Other tasks currently bundled into a driver’s job, like securing cargo and refueling, would still need a human at some point in the chain. The net effect on total employment is genuinely uncertain, but the nature of trucking work will change even for drivers who keep their jobs, with shorter regional routes replacing cross-country hauls.
A Realistic Timeline
Autonomous trucks are already moving freight commercially on limited routes in the Sun Belt. Expansion beyond those corridors will depend on three things: sensor technology improving enough to handle rain, snow, and poor visibility; regulators creating clear rules for driverless commercial vehicles; and the economics working out at scale, including the cost of transfer hubs, remote monitoring staff, and specialized maintenance.
The most likely trajectory is a slow geographic expansion outward from the southwest, following warm-weather interstate corridors first and reaching northern and mountainous routes last, if at all in the near term. Full automation of the entire U.S. trucking network, where no human drivers are needed anywhere in the supply chain, is not on any realistic horizon. What’s happening instead is a restructuring: autonomous systems taking over the highway segments they do well, with human drivers handling everything else.