Autonomous Driving May Be Nearing Its ‘ChatGPT Moment’
A T. Rowe Price report projects that robotaxis and self-driving trucks could scale commercially within three to five years, with a turning point from 2027.
Autonomous driving may be nearing a major commercial turning point, according to a T. Rowe Price report that projects robotaxis and self-driving trucks could scale within three to five years. The authors expect the shift to become clearer from 2027 onward.
David Giroux, chief investment officer of T. Rowe Price Investment Management, co-wrote the report with portfolio manager Mike Signore. They described autonomous vehicles as a potential large-scale use of artificial intelligence after about 15 years of delays and failed projects. Ford Motor Co. and Volkswagen closed their Argo AI venture in 2022.
Self-driving trucks could expand faster than robotaxis because long-haul freight uses established routes and is subject to federal limits on driver hours. Truck drivers may drive for up to 11 hours during a 14-hour shift before taking 10 hours of rest. A 17-hour trip from Dallas to Phoenix could therefore take two days with a human driver, while a driverless truck could potentially complete it in one day.
Aurora Innovation expects to operate more than 200 driverless heavy-duty trucks by the end of 2026. Its targets are more than 1,000 trucks by the end of 2027 and 30,000 by 2030. Volvo estimated in June that nearly one in 10 U.S. long-haul trucks could be autonomous by 2035.
Alphabet-backed Waymo operates commercial driverless robotaxi fleets in 15 U.S. cities, according to the report. The company plans to expand to 17 additional locations, including London and Tokyo.
Safety remains a requirement for wider commercial deployment. Government and academic research cited in the report indicates that driver-related factors, including human error, contribute to more than 90% of serious crashes.
“Driverless technology that is 98% of the way there is still very far from being ready for commercial operation,” Giroux and Signore wrote.
The report contrasts Tesla’s camera-only system with the multi-sensor systems used by Waymo and Aurora. Cameras can reduce costs, but the authors noted that they may perform less reliably in poor weather and at higher speeds. They also described a single end-to-end artificial intelligence model as harder to diagnose when it fails.
Waymo and Aurora combine cameras with radar and LiDAR, a technology that uses laser pulses to map nearby objects. The report projects that radar and LiDAR costs will decline over time, similar to the fall in automotive camera costs during the 2000s.
The authors described a 200-mile intercity autonomous-vehicle ride during which the vehicle changed lanes at 70 miles per hour without an obvious reason. A police cruiser and a stopped vehicle appeared several seconds later. The LiDAR detected the vehicles before they were fully visible to the passengers.
T. Rowe Price’s Innovation Leaders ETF, known by its ticker TNXT, holds both Alphabet and Tesla. As of Oct. 28, Alphabet represented 1.95% of the fund and Tesla represented 1.80%. The actively managed ETF launched in January and charges an annual expense ratio of 0.49%.
The fund’s net asset value rose 14.6% from its launch through Sept. 30, compared with a 9.9% return for the Russell 3000 Index, according to T. Rowe Price. The report compares the potential structure of the autonomous-driving market with food delivery and ride-hailing, where a small number of companies hold most of the market.








