BMW, Volkswagen and Mercedes-Benz keep announcing driverless fleets and keep running into the same walls. China, meanwhile, already has them running around the clock.
Plenty of carmakers have promised robotaxis over the past few years. Most of those promises have quietly evaporated, with companies reversing course not long after the press release went out.
Tesla, the loudest voice in the category, has seen its own rollout slow over recent months under pressure to tighten safety and with operational and software problems still unsolved. It has company. Volkswagen, BMW and Mercedes-Benz are all hitting serious obstacles with their own robotaxi ambitions.
Slow, uneven progress
Legacy manufacturers have made some headway, but nothing close to what a wide commercial launch would require.
Volkswagen set up its autonomous driving subsidiary, VW ADMT, in July 2023, and announced a commercial robotaxi partnership with Uber in April 2025. This July, it finally got a real pilot passenger service on the road in Hamburg through its Moia subsidiary.
Getting from that pilot to full commercial operation across Europe is another matter. Regulatory timelines are tight, hardware is expensive, and safety monitors are mandatory. European and German road rules require a trained monitor behind the wheel during public testing, which makes fully driverless commercial certification for Volkswagen unlikely before 2027 at the earliest.
Europe is also warier of unproven software than the US or China. That caution delays the real-world training algorithms need. Local safety watchdogs and crash-investigation procedures may not be able to move at the speed a fast commercial rollout would demand either.
The money problem
Costs have done as much damage as regulation. BMW has paused the “eyes off” Personal Pilot Level 3 features on its flagship models, citing rising costs and weak demand, and redirected effort into Level 2 Plus and DCAS assisted driving, where the driver stays responsible for the car and has to keep watching the road. Volkswagen, for similar cost reasons, walked away from a joint development alliance with Bosch.
Buyers are not helping the business case. Fewer are willing to pay a premium for limited autonomous features, which leaves driverless fleets looking shaky on paper. Competition between European players such as Mercedes-Benz and tech partnerships like the Uber-Autobrains/Wayve tie-up has squeezed the assisted and automated driving market further.
There is a structural issue too. European cities already have strong public transport, so attention has drifted towards integrating with transit rather than adding more ride-hailing.
What China is doing differently
China already runs robotaxis in Beijing, Wuhan, Shenzhen, Guangzhou and Shanghai. Services mostly operate inside geofenced pilot zones and suburban districts rather than across whole metropolitan areas, though Wuhan and Shenzhen have large commercial fleets running 24/7. Riders hail them through platforms they already use, such as WeChat and Alipay.
Government backing is a big part of it, with state policy supporting autonomous driving as part of national technology goals. Safety incidents happen there too. A technical traffic disruption in Wuhan in April this year triggered a licensing pause, but permit issuance has been recovering since, as of July. Setbacks slow individual cities without derailing the wider push.
China’s car manufacturing base and EV supply chain make hardware cheaper and production easier to scale. Its cities offer another advantage: driving through the traffic in Beijing or Guangzhou teaches software to handle pedestrians and vehicles that refuse to behave predictably. And Chinese passengers have shown more willingness to get into a driverless car than riders in Europe and other Western markets.
