Stellantis has also reached a preliminary agreement with Pony.ai to jointly produce robotaxis for European cities, combining the automaker’s electric van architecture with a self-driving software stack that the firm describes as mature. The strategy hinges on eventual integration of Pony.ai’s self-driving technology into Stellantis’ mid-sized electric vehicle platform, building new vehicles from scratch at the factory that will be purpose-built for autonomous driving.
The first prototypes will be developed on the Peugeot e-Traveller and testing is scheduled to start in Luxembourg, where Pony.ai is permitted, before rolling out in further urban markets. The partnership focuses on geofenced services that can scale up in a phased operation, as safety, regulatory approvals and operational readiness come together.
Why this Stellantis–Pony.ai pairing matters for Europe
As autonomy shifts revenue from car sales to miles delivered, automakers are more and more looking to keep value in-house. Stellantis provides the muscle of manufacturing scale across a number of brands and an updated lineup of mid-vans, while Pony.ai is bringing to the table a Level 4 autonomy stack that has been quite perfected over years of robotaxi pilots and fleet operations in tricky Asian megacities. It’s a textbook “hardware meets software” equation, the hardware being sold now with a cut of recurring revenue from autonomous mobility services in the future.
Pony.ai’s recent partnership with Uber to research driverless ride-hailing in global markets suggests a channel for distribution when vehicles are ready for commercial deployment. For Stellantis, the effort aligns with its software strategy and mobility ecosystem more broadly, offering up its vans as customizable platforms for delivery, shuttle and ride-hail applications that can be updated over the air.
What will be on the road in the first robotaxi rollout
The e-Traveller–based robotaxi will be built on Stellantis’s newest electric van architecture, which promises impressive urban range and crucial electrical headroom for the autonomous hardware.
Existing models in the family provide more than 300 km of WLTP range, and the underpinnings support 400-volt wired fast charging, a valuable feature when you have to turn rental cars around in a hurry.
Pony.ai generally uses a sensor fusion suite that combines lidar, radar and high-definition cameras built on top of redundant compute and safety-critical actuation. Look for drive-by-wire steering and braking, isolated power domains and data recorders to help support incident analysis. The autonomy should take place in a specified Operational Design Domain, first focused on mapped city areas with well-managed curbs and predictable traffic flow.
And since European cities are famously dense and well-peopled, the service design will matter as much as the autonomy stack. Look for staged rollouts during off-peak times, remote assistance to get people out of trouble at the edges and accessible features that make the vans usable for wheelchair users and riders with limited mobility — an area where electric multi-row people-carriers can truly shine.
Europe’s regulatory reality check for autonomous robotaxis
Europe’s journey toward driverless services winds through a layered regulatory regime. Type-approval regulations under the United Nations Economic Commission for Europe apply to fundamental technologies such as automated lane keeping (UN R157), cybersecurity systems (UN R155) and software updates (UN R156). In addition, pilot operations will have to be cleared by individual countries for public roads, and liability and insurance frameworks will also have to be clarified.
Luxembourg has granted Pony.ai approval to test and thus a practical point of departure. Nearby markets like France and Germany have also developed legal frameworks for more automated pilots under stringent operating circumstances. Compliance will necessitate a robust safety case that covers ISO 26262 functional safety, ISO 21448 (SOTIF) with respect to edge cases, and increasingly UL 4600 for autonomous system assurance. Data governance is another challenge, one that will be colored by GDPR and how fleet operators record, store and process sensor data.
Competitive landscape and economics of European robotaxis
Despite aggressive pilots by global players, Europe has not seen widespread commercial robotaxi deployment. Mobileye has shown it can go driverless with partners in test cities, and UK developers, such as Oxa and Wayve, are pursuing similar methods to urban autonomy. Enter Stellantis and Pony.ai, wagering that a factory-integrated, multi-brand fleet can move faster than custom-built retrofits.
Its business case is strong, if it can drive its cost per kilometer below that of human-driven ride-hailing. Consultancy industry analyses predict the European market for autonomous mobility services will be worth multiple billions of euros over the next decade, with unit economics depending on high uptime, durable sensor hardware and efficient depot charging. Research from the International Transport Forum, in fact, estimates that shared autonomous offerings could drastically cut both private car dependence and congestion—arguments that urban leaders increasingly want to hear when deciding whether or not to grant permits for driverless fleets.
What to watch next as Stellantis and Pony.ai scale pilots
Three cues will decide if this partnership scales.
- Safety performance in early pilots: Disengagements are not standardized; seek independent safety audits, transparency of incidents and consistent operation across varying weather and lighting.
- Speed of regulatory approvals beyond the initial test market: Formal exemptions, insurance plans and labor considerations won’t help if they’re not in place to conform to a timeline.
- Fleet readiness: Strong lidar and compute supply chains, serviceable sensor placements and OTA update pipelines that adhere to cybersecurity standards.
Assuming Stellantis can standardize an “autonomy-ready” van setup and that Pony.ai can plug itself into ride-hailing partners for demand, Europe’s first impactful wave of robotaxis might not come as a one-off pilot but as a replicable service model.
