The self-driving truck is coming faster than

There’s a $700 billion industry in the United States that moves 72% of all freight by weight, and it can’t find enough humans to do the work.

The trucking industry has had a driver shortage for years. The American Trucking Associations estimates the shortage at 80,000 drivers and growing. The work is hard, the hours are long, the lifestyle is punishing. Young people don’t want to drive trucks. The average age of a truck driver keeps rising.

Meanwhile, Aurora, Kodiak Robotics, and others are running autonomous trucks on real highways in Texas and the Southwest. Not test tracks. Real highways, with real freight, sharing lanes with regular traffic.

I think autonomous trucks will be mainstream before autonomous taxis.

Why trucks are easier than cars

This sounds counterintuitive. Trucks are bigger, heavier, harder to stop. But the driving task is simpler.

Highway driving is structured. Lanes are wide and well-marked. Speeds are relatively constant. There are no pedestrians, no cyclists, no dogs running across the road. The primary challenges are lane keeping, following distance, and handling merges.

City driving is chaos. Pedestrians jaywalk. Cyclists appear from blind spots. Double-parked cars create obstacles. Traffic lights fail. Construction zones change daily. Children chase balls into streets.

Waymo has spent years and billions of dollars mastering city driving. The autonomous truck companies get to skip most of that complexity. Highway miles are what matter for freight, and highway miles are the easy miles.

The economics

A human truck driver costs about $70,000-80,000 per year. But the real cost is the constraints. A human driver can only legally drive 11 hours per day. A truck that’s not moving is a truck that’s not earning. The utilization rate of a human-driven truck is roughly 45%.

An autonomous truck can drive 20+ hours per day (stopping only for fuel and maintenance). The utilization rate more than doubles. A fleet of autonomous trucks can move the same freight with fewer vehicles, fewer stops, and more predictable schedules.

The fuel savings matter too. Autonomous trucks can maintain optimal speed, avoid unnecessary braking, and platoon (drive close together in a line to reduce aerodynamic drag). Estimates suggest 10-15% fuel savings from optimized driving alone.

For a $700 billion industry, even modest efficiency gains translate to tens of billions in savings. The economic incentive to automate trucking is massive, immediate, and growing with every undelivered shipment.

The approach

Most autonomous truck companies use a “hub-to-hub” model. The truck drives autonomously on the highway between distribution hubs. At each hub, a human driver takes over for the complex local driving: backing into loading docks, navigating warehouse parking lots, handling city streets.

This is pragmatic and probably right. It solves the easy problem first (highway driving), captures most of the economic value (long-haul miles are where the money is), and avoids the hard problem (city driving) entirely.

Over time, the “city driving” portion shrinks as autonomous capability improves. But the hub-to-hub model works today. It doesn’t need full autonomy. It needs highway autonomy. And highway autonomy is close.

What this means for drivers

The question everyone asks. The honest answer: long-haul trucking jobs will decline. Not disappear overnight, but decline. The 80,000 driver shortage provides a buffer. Autonomous trucks fill the gap that can’t be filled by humans first.

But driving jobs won’t go to zero. Local delivery, complex urban routing, specialized cargo that needs human judgment, these will remain human-driven for a long time. The trucker of 2035 might be someone who drives the last 50 miles instead of the middle 500.

That’s a different job. Whether it’s better or worse depends on who you ask. A long-haul driver who loved the open road will lose something they can’t get back. A local driver who gets to sleep in their own bed every night might see it as an upgrade.

The transition will be messy, uneven, and politically charged. It always is when machines take over human work. The trucking industry knows it’s coming. The unions know. The truck stops know. Everyone’s adjusting, just at different speeds.

I think autonomous trucks will be carrying freight on major US highway corridors within 3 years. Maybe sooner. The technology is there. The economics are overwhelming. And the trucks don’t need a driver who can’t be found.

Related thinking: