Heavy-Duty AVs Gain a Commercial Path, Approved Corridors Become the Moat, and Permitted Miles Define Robotaxi Credibility

Heavy-Duty AVs Get a Real Commercial Path in California

California’s April 28 DMV update did more than lift the 10,001-pound GVWR testing ban: it created a phased path for heavy-duty commercial AVs to move from safety-driver testing to driverless trials and, eventually, deployment. That widens the commercialization lane beyond passenger robotaxis just as other approvals are expanding service footprints. Waymo won CPUC approval for driverless passenger service across Sacramento and Yolo counties, while Nevada authorized what is being described as the largest U.S. robotaxi rollout yet: up to 8,000 vehicles across Clark County over the next 12 months, including as many as 5,000 from Tesla, 1,000 from Waymo, and 1,000 from Uber. Tesla and Baidu expanded driverless operations in key cities; Baidu and Uber pushed further globally; Uber and Pony.ai debuted robotaxis in Europe. Waymo’s push to lower the cost of its Ojai fleet and partnerships such as Aebi Schmidt-Yeti Move and Nissan/Honda’s alliances with AV startups show that cost discipline and capability access are becoming as important as autonomy performance. The progression from regional permissioning to repeatable fleet expansion is now visible across vehicle classes, and practitioners should focus on fleet integration, compliance execution, and lower-cost deployment systems that can scale across markets.

Approved Corridors and Fleet Operations Are Becoming the AV Moat

Pony.ai’s cross-provincial robotruck platooning on the Beijing–Tianjin–Hebei corridor and Aurora’s push toward 200 driverless trucks by year-end 2026 show AV commercialization shifting from demos to regulated routes, contracted freight, and fleet scale. In China, Pony.ai is operating on approved roads under the “1+N” model, with one human-driven lead truck and autonomous followers on routes including the Beijing–Tianjin–Tanggu and Jingjintang expressways; its freight service between Beijing and Tianjin has reportedly run with Sinotrans/Sinotrans Logistics.

Aurora’s case shows the same economics in the U.S. The company says it has launched second-generation driverless trucking on public roads, logged more than 250,000 driverless miles by January 2026, secured 10 approved driverless routes and 12 active lanes, and expects Roush to ramp manufacturing to about 20 trucks per week by October. The competitive advantage is moving to corridor access, fleet utilization, and manufacturing throughput, not autonomy validation alone.

Permitted Miles Are Emerging as the Robotaxi Moat

Tesla’s Austin robotaxi program is now competing on regulator-usable operating evidence, not just autonomy claims: in the mid-June to mid-July 2026 reporting window, Tesla logged two robotaxi incidents in NHTSA submissions and described both as not at fault, while the Robotaxi Tracker recorded 170 unsupervised rides across 54 vehicles over the same two weeks. That matters because unsupervised miles plus favorable incident classification increasingly determine who can expand service, support underwriting, and argue commercial readiness, even as NHTSA does not independently assign fault from Tesla’s filings. Seoul is following the same logic, expanding its Gangnam robotaxi fleet from 7 to 19 vehicles and widening the 20.4 km² pilot area with approved safety plans, reduced speeds, alerts, and a human safety operator. The progression from safety proof to priced risk is now becoming a contest over permitted operating miles and the quality of the evidence behind them. Operators that can document those miles cleanly will be better positioned for expansion, insurance, and regulatory approval; those that cannot will keep running into the same commercial bottlenecks in a more formalized way.