Autonomous trucking uses self-driving technology to move freight on highways with limited or no human input. The industry is in its earliest commercial stages, with companies like Aurora Innovation running trucks on specific highway corridors, though significant technical and regulatory hurdles remain before driverless semis become a common sight on American roads.
How the Automation Levels Work
Not all autonomous trucks are created equal. The National Highway Traffic Safety Administration classifies vehicle automation on a scale from Level 1 to Level 5, and understanding where a truck falls on that scale tells you a lot about what it can and can’t do.
At Level 1, a system handles either steering or speed control, but not both. Adaptive cruise control on a modern semi is a good example. Level 2 adds both steering and speed control working together, but the driver is still fully responsible and must stay engaged at all times. Most of what you’d see in newer commercial trucks today falls into these two categories.
Level 3 is where things shift meaningfully. The system handles all driving tasks, but a human must be available in the cab to take over when the system requests it. Think of this as “the truck drives, you’re on standby.” Level 4 takes another leap: the system is fully responsible for driving within a limited area or set of conditions, and no human driver is needed during those stretches. Level 5 means the system drives everywhere, in all conditions, with no human involvement required at all. Level 5 trucking does not exist yet, and no company is close to deploying it.
The commercial activity generating the most attention right now is happening at Level 4, where trucks can drive themselves on defined highway corridors but still need human involvement for complex situations like navigating city streets or construction zones.
Where Autonomous Trucks Are Running Today
Aurora Innovation launched what it called the first fully autonomous commercial trucking operation on U.S. highways, running Peterbilt trucks between Dallas and Houston along the I-45 corridor. Those runs hauled freight for customers including FedEx and Uber Freight and quickly logged 1,200 miles. But the rollout wasn’t seamless. Less than three weeks after starting, the runs paused after Paccar, the manufacturer of the Peterbilt trucks, asked Aurora to put a person back in the cab. Operations resumed with an “observer” on board, someone Aurora describes as less engaged than a traditional safety driver but still present.
That sequence captures the current reality of autonomous trucking: the technology can handle stretches of highway driving, but the industry is still working through questions about safety oversight, manufacturer liability, and how quickly to remove humans from the equation. Other companies are testing autonomous trucks on highway corridors as well, generally in sunbelt states where weather conditions are more predictable and regulatory environments have been more welcoming.
Why the Freight Industry Is Interested
The economic case for autonomous trucking centers on cost reduction and capacity. Driver wages and benefits represent one of the largest expenses in long-haul trucking, and the industry has faced a persistent shortage of drivers willing to spend weeks away from home. An autonomous truck doesn’t need rest breaks mandated by hours-of-service rules, meaning it could theoretically run nearly around the clock, cutting delivery times on long routes.
A freight modeling study from the U.S. Department of Transportation examined what would happen if autonomous trucking costs dropped to half those of human-driven trucks. The model predicted truck mode share would rise by 4.2 percentage points, from 57% to 61.2%, with a 6% increase in ton-miles transported by truck. In other words, cheaper autonomous freight wouldn’t just replace existing truck runs. It would pull cargo away from rail and other modes of transport, expanding trucking’s share of the overall freight market.
That shift is still theoretical, but it explains why major logistics companies and truck manufacturers are investing heavily. Even modest cost reductions on high-volume freight corridors could reshape how goods move across the country.
Weather and Sensor Limitations
The biggest technical barrier to widespread autonomous trucking isn’t highway driving on a clear day. It’s everything else. Autonomous trucks rely on a combination of sensors, and each type has significant weaknesses in bad weather.
LiDAR sensors create detailed 3D maps of the truck’s surroundings using laser pulses. They work extremely well in clear conditions but struggle when roads are covered in snow, which obscures lane markings and road edges. Cameras, which handle tasks like reading signs and identifying lane lines, are degraded by fog, heavy rain, snow, and dust. Even sunny conditions can cause glare that compromises camera performance. Radar is the most weather-resistant sensor, capable of detecting large moving objects through fog, rain, and dust, but it can’t operate independently to provide a complete picture of the driving environment.
Ultrasonic sensors, used for close-range object detection, can be triggered by falling snow, creating continuous false alarms. Wet and icy surfaces cause reflectivity problems that throw off sensor readings across the board. And the communication systems that allow trucks to share data with other vehicles and road infrastructure can be disrupted by fog, heavy rain, hail, and thunderstorms.
Beyond sensor performance, slippery roads require the truck to reduce speed and increase following distance, something that demands complex algorithms to detect conditions and adjust behavior in real time. Heavy winds and hail add another layer of difficulty. The Department of Transportation has noted that technical standards for how autonomous vehicles should perform in adverse weather don’t fully exist yet, with efforts still underway to define those requirements.
The Hub-to-Hub Model
Rather than trying to build a truck that can handle every driving scenario, most companies are pursuing a “hub-to-hub” or “middle mile” approach. The autonomous truck handles the highway portion of a route between two transfer facilities, while human drivers manage the first and last miles through city streets, loading docks, and complex local roads.
This approach plays to the technology’s strengths. Highway driving is more structured and predictable than urban driving. Lanes are wider, speeds are more consistent, pedestrians and cyclists are absent, and intersections are replaced by on-ramps and off-ramps. By limiting the autonomous system to this environment, companies can deploy trucks commercially while the technology for more complex driving continues to develop.
At the transfer hubs, a human driver picks up the trailer and completes the delivery. This means autonomous trucking in its current form doesn’t eliminate trucking jobs entirely. It changes them, potentially shifting long-haul positions into shorter local routes that let drivers return home each night.
What This Means for Trucking Jobs
The trucking industry employs roughly 3.5 million drivers in the United States, and questions about job displacement are central to the autonomous trucking conversation. The near-term reality is more nuanced than “robots replace drivers.” The hub-to-hub model still requires humans for local driving, and the technology is initially targeting the long-haul routes that are hardest to staff. The industry has reported driver shortages for years, particularly for over-the-road positions that keep drivers away from home for extended periods.
New roles are also emerging around autonomous trucking operations. Companies need remote monitors who oversee fleets of autonomous trucks, technicians who maintain sensor systems, and operators at transfer hubs. The transition, if it happens at scale, would likely unfold over years or decades rather than overnight, given the pace of regulatory approval and the capital investment required to replace conventional fleets.
Regulatory Landscape
There is no single federal law that permits or prohibits autonomous trucks on U.S. highways. Regulation has developed as a patchwork, with some states passing laws that explicitly allow autonomous vehicle testing and commercial operations, while others have no specific framework in place. The Federal Motor Carrier Safety Administration oversees trucking safety nationally, but many of its existing rules, like hours-of-service requirements, were written with human drivers in mind and don’t translate neatly to driverless operations.
Insurance and liability questions remain largely unresolved. When an autonomous truck is involved in a crash, the question of who bears responsibility shifts from the driver to a web of potentially liable parties: the software developer, the truck manufacturer, the sensor supplier, and the fleet operator. These questions will likely be settled through a combination of state legislation, federal rulemaking, and court cases as the technology moves from pilot programs to broader deployment.