The $600 Handoff: Why I’m Calling a U-Turn on Hub-to-Hub     

I’m calling it: I was wrong.

A year ago, I was convinced — and I told anyone who’d listen — that the hub-to-hub model was the only way forward for autonomous trucking. It was the “Golden Bridge” to the future. It looked clean on a PowerPoint. The robot handles the so-called boring 500 miles on the I-10, and a human driver handles the messy city streets. We’d meet at a transfer hub at the edge of town, swap trailers like a relay race, and everyone would win.

But as we hit April 2026, the industry’s heavyweights are calling an audible, and for good reason.

The transfer hub dream is hitting a wall, and the math has finally caught up to the hype. If you’re still planning your fleet’s future around a mid-journey handoff, you’re planning for a $600 per load deficit that will eat your ROI alive. This isn’t just a preference change; it’s a fundamental shift in the economic physics of freight.

The Math Has Changed: Why the “Big Two” are Walking Away

It’s not just me. The two biggest players in the game, Aurora Innovation and Waabi, have essentially spent the first quarter of 2026 admitting that the hub-to-hub model is an economic dead end for the long haul.

Lior Ron, CEO of Waabi, recently called transfer hubs crutches that don’t scale. He’s right. Waabi’s shift toward direct-to-customer (D2C) navigation acknowledges a truth we tried to ignore: 90% of freight doesn’t want to stop at the edge of town. It wants to go to the dock.

Chris Urmson, co-founder of Aurora, doubled down on this at CES 2026. Aurora has shifted its entire mapping and deployment strategy toward customer endpoints. Why? Because when you sit down with a carrier and look at the P&L, the $600 “handoff tax” at a transfer hub erases every penny of the margin the robot just saved you on the highway.

The “Handoff Tax” That Kills the Margin

We’ve spent years perfecting the “brain” of the truck for the highway, but we ignored the friction of the handoff. When an autonomous truck pulls into a transfer hub, the clock doesn’t just stop; it starts draining cash. Let’s break down that $600 “leak.”

1. The drayage tax ($300 to $450): You still have to pay a human driver to pull the trailer from the customer dock to your hub. In a tight labor market, the local short haul is the most expensive mile in the system.
2. The real estate albatross ($150): Someone has to pay for that 20-acre lot, the fencing, the security, and the lighting. These aren’t just dirt lots anymore; they are high-liability transition zones.
3. Dwell time ($100+): Every minute an autonomous power unit sits idle waiting for a human driver to show up with a trailer is a minute that a $250,000 asset isn’t depreciating against a revenue mile.

Total it up, and you’re looking at a massive friction cost. In a world where an AV was supposed to save you $0.35 per mile, that $600 tax erases the savings of a 1,700-mile haul. If your average length of haul is 800 miles, the hub model actually makes you less profitable than a traditional diesel team.

The Dock Dance: Why the Hard Ends are Hard

If the highway is solved, the new frontier is the messy middle — the uncontrolled environment of the customer loading dock. We used to think this was impossible for AI, which is why we built the hubs. But the technology has matured.

The dock dance is the ultimate test of an AV’s perception system. On the highway, you have lane lines and predictable traffic flow. At a DHL or Walmart distribution center at 4 a.m., you have occlusion, or objects blocking other objects. You have a forklift driver darting out from between parked trailers. You have steam from a cooling unit creating noise in the LiDAR point cloud.

The industry is realizing that we can’t hide the robots in safe zones anymore. To win, the AV brain has to navigate the Wild West of the customer site. This is why mapping endpoints is the new gold rush. If the truck can’t back into Bay 42 without a human taking the wheel, the business model collapses.

The SELF DRIVE Act: Who Grades the Code?

The SELF DRIVE Act (H.R. 7390) is the defining regulatory crossroads of 2026. As we pivot to D2C delivery, the bill highlights a fundamental friction point between the people building the technology and the people buying it.

On one side, OEMs need Federal Preemption. They argue, rightly so, that a 50-state patchwork of safety laws would kill interstate autonomous freight before it starts. They need a single national standard that allows software to update at the speed of AI, not at the speed of government bureaucracy.

On the other side, fleets are raising the alarm on self-certification. In its March 9, 2026, letter, OOIDA called the current bill “homework without a teacher.” The legislation allows manufacturers to submit Safety Cases only upon request, meaning the public often won’t know if a system is safe until after a failure occurs.

For a carrier, this creates a massive verification gap. If you skip the safe highway hub and send a robot into the Wild West of a customer’s loading dock, you are betting your company’s survival on a software patch that hasn’t been independently verified. The industry doesn’t need to choose sides; it needs transparency. We need a framework where OEMs can innovate at scale, but fleets aren’t left holding the primary liability for “self-certified” code that isn’t ready for a 4 a.m. 90-degree back-in.

The Life Raft for the Mid-Size Carrier

There is a silver lining here, specifically for the mid-size carriers. For a long time, the hub-to-hub model favored the mega-fleets that could afford to build massive real estate networks. It threatened to turn everyone else into a drayage provider for the robots, picking up the scraps of the first and last mile.

By pivoting to D2C, the playing field levels out. If the truck can go dock-to-dock, the infrastructure becomes the existing highway and the existing customer dock. You don’t need a 50-hub network to compete. You just need a truck that is smart enough to do the job from start to finish. This shift is a life raft for carriers who were worried about being blocked out of existence.

Dean’s Bottom Line

The U-turn is here. We are moving from a technology-first era (where we built hubs because the robots were scared of cities) to an economic-first era (where we go direct-to-customer because the hubs are too expensive).

Autonomy isn’t a highway-only game anymore. It’s a dock-to-dock game. The easy middle is solved. We know how to keep a truck in a lane at 65 MPH. But the hard ends are where the profit lives, and where the legal risks are peaking.

The SELF DRIVE Act might give manufacturers the green light to skip the government line, but it doesn’t give you a pass on the liability when things go sideways at the warehouse gate. If you’re a fleet manager, stop asking where the nearest hub is. Start asking your AV partners how their sensors handle a tight 90-degree back-in at your busiest customer site.

The $600 handoff is a luxury the freight industry can’t afford. It’s time to go door-to-door. The question is simple: Are you building for the hub, or are you building for the dock? Because the industry just picked a side, and the ROI is waiting at the loading bay.

Feature image generated with Google Gemini.