The most expensive part of a robotaxi is the car. The most expensive part of a car-sharing fleet is the driver. Niulinx, a spin-off from Politecnico di Milano, has just raised €38 million to see if it can build a business by eliminating the latter and retrofitting the former [MarketScreener, April 2026]. The bet is that a full-stack autonomous driving system, bolted onto existing vehicles and limited to low-speed urban routes, is a capital-efficient wedge into the future of shared mobility. Their first road tests are slated for the streets of Milan soon [Il Sole 24 ORE, 2026].
A consortium of Italian incumbents
What makes Niulinx's €38 million pre-seed round notable is not just its size, but who wrote the checks. The investor list reads like a roll call of Italy's industrial and public infrastructure backbone [MarketScreener, April 2026].
| Investor | Type | Strategic Angle |
|---|---|---|
| A2A | Multi-utility | Urban energy grids, smart city infrastructure |
| Gruppo Ferrovie dello Stato Italiane | National railway operator | First/last-mile transport, mobility-as-a-service |
| Pirelli | Tire manufacturer | Vehicle data, tire-as-a-sensor integration |
| CDP Venture Capital | State-backed VC | National tech sovereignty, deep tech |
| MOST - National Center for Sustainable Mobility | Research consortium | Public R&D, pilot program access |
This isn't venture capital chasing a moonshot. It's a consortium of strategic players placing a coordinated bet on a specific, asset-light path to autonomy. The presence of the national railway and a major utility suggests the initial customer isn't a consumer, but a municipality or a public transport operator looking to extend service. As one Italian paper put it, Niulinx represents "the Italian challenge to robotaxis" [Il Sole 24 ORE, 2026].
The retrofit wedge
The company's technical premise is straightforward, which is its main appeal. Instead of designing a bespoke, sensor-laden vehicle from the ground up,a path that has consumed billions,Niulinx is developing a stack to convert standard production cars. The system handles environmental perception, route planning, vehicle control, and remote fleet management [Startup.eu]. The "low-speed" qualifier is crucial. It narrows the operational design domain to complex but manageable urban environments, like dedicated lanes or geo-fenced districts, where speeds are low and the cost of a mistake is more manageable.
From a climate and energy perspective, this wedge has a certain logic. The embodied carbon in a new vehicle is significant. A retrofit model, applied to an existing car-sharing fleet, could theoretically increase asset utilization without requiring a whole new generation of manufactured hardware. The unit economics start with the price of the sensor kit and the software subscription, not the price of a new car.
The academic engine room
Niulinx was founded in 2025 as a spin-off from the Department of Electronics, Information and Bioengineering (DEIB) at Politecnico di Milano, specifically from the AIDA research group led by Professor Sergio Matteo Savaresi [Politecnico di Milano]. The founding team is large, with nine individuals listed, suggesting a deep research bench. CEO Luca Foresti brings experience from scaling a healthtech company, while CFO Gian Marco Felice adds financial governance from his role at ProCredit Holding AG [Economyup, 2026] [Bloomberg Markets, 2026].
The academic origin provides credibility in the core algorithms but leaves a key question unanswered. Building the stack is one discipline. Deploying, maintaining, and operating a commercial robotaxi service at city scale is another, requiring deep expertise in safety validation, municipal regulation, and fleet operations that exists outside the lab.
Where the wheels could come off
The ambition is clear, the funding is substantial, and the backing is strategically astute. The risks, however, are just as concrete.
- The commercialization gap. The company is pre-deployment. Road tests in Lombardy are imminent, but moving from controlled testing to a revenue-generating service with public passengers involves a gulf of regulatory, insurance, and public acceptance hurdles that have stalled far more resourced players.
- The speed limit. Focusing on low-speed environments simplifies the technical challenge but also limits the addressable market and potential revenue per vehicle-hour. It confines the service to a niche within the broader autonomous mobility landscape.
- The global context. Niulinx is entering a field where giants like Baidu's Apollo Go have logged millions of rides and where capital has flooded,and occasionally retreated,for over a decade. Being a capital-efficient, retrofit-focused European player is a differentiation, but it doesn't exempt them from solving the same core problems of safety and scalability.
The consortium of investors is arguably Niulinx's most potent early asset. A2A and Ferrovie dello Stato aren't just financiers; they are potential launch customers and partners with the operational heft to navigate Italian bureaucracy and pilot projects.
The next twelve months
The immediate milestone is the start of on-road testing in Italy. Success won't be measured by miles driven, but by a clear transition from a research project to a defined commercial pilot with a named municipal or transport partner. The €38 million war chest means they don't need to raise again immediately, giving them runway to focus on this technical and regulatory proving ground.
A back-of-the-envelope calculation puts their initial advantage in perspective. If a purpose-built robotaxi from a major player costs over $200,000 (estimated), Niulinx's retrofit kit might target a fraction of that. Their path to unit economics hinges on that delta being large enough to cover the costs of retrofitting, operations, and the inevitable regulatory overhead, while still undercutting the total cost of a human driver in a car-sharing scheme.
For Niulinx to succeed, it doesn't need to beat Waymo in San Francisco. It needs to out-execute the traditional bus and shuttle contractors in Milan, proving that a retrofitted, autonomous fleet is a safer, more reliable, and ultimately cheaper way to move people through the last mile of an Italian city. That's a narrower race, but a real one.