Aurora's driverless trucks go commercial for McLane in Texas
Aurora Innovation just inked a commercial deal with McLane, one of the largest distribution companies in the US. Starting now, Aurora's self-driving trucks will haul freight between Dallas and Houston without a human safety driver behind the wheel. This is not a pilot — it's a live, paying contract.
The deal in detail
The trucks operate autonomously for the long-haul portion of the route. When they reach a terminal just off the highway in Dallas or Houston, a McLane driver takes over for local deliveries to restaurants and stores. This handoff model keeps autonomous tech on the easy part (highway) and lets humans handle complex last-mile stops.
Aurora still has a "human observer" in the cab — not a driver, but someone who monitors the system. That's because of an agreement with truck manufacturer Paccar, not because the tech needs it. Aurora's system is capable of Level 4 autonomy (no human intervention required) on these routes.
From pilot to production
This didn't happen overnight. Aurora and McLane started a pilot in 2023 with safety drivers. They gradually expanded to two round-trips daily between Dallas and Houston. Now, the route runs seven days a week with no safety driver. The commercial contract is a direct result of that pilot's success.
What this means for Aurora
Aurora has been developing autonomous truck tech for years. This deal is a clear signal they're moving from R&D to revenue. They already have a contract with Detmar Logistics for frac sand hauling, and Hirschbach Motor Lines agreed to buy 500 Aurora-powered trucks (though that's still a memorandum of understanding).
Today, Aurora operates driverless trucks on several Texas routes: Dallas-Houston, Fort Worth-El Paso, El Paso-Phoenix, Fort Worth-Phoenix, and Laredo-Dallas. Some still have observers; others are fully autonomous. The company plans to expand to new McLane distribution centers across the US Sun Belt by the end of 2026.
Why developers should care
This is a real-world deployment of AI in a safety-critical system. The stack includes perception, planning, control, and V2X communication. For ML engineers, it's a case study in taking a model from simulation to production with rigorous validation. For systems engineers, it's about building fault-tolerant architectures that can handle highway speeds and 80,000-pound vehicles.
The handoff model is also interesting: autonomous for the easy part, human for the hard part. That's a pragmatic approach to incremental deployment that many teams could learn from.
The bottom line
Aurora is now a commercial operator, not just a developer. The McLane deal is revenue, not hype. If you're building autonomous systems, watch how they handle the transition from pilot to paid service — that's where the real engineering challenges live.
