Who does carsharing better – OEMs or start-ups?

For a taxi fleet owner, autonomous vehicles may seem to remove the most important part of the current business – the driver. But even without drivers, someone still needs to take care of the cars. They need to be owned or financed, charged, cleaned, inspected and repaired. Someone also needs to manage depots, deal with incidents, help customers and make sure enough cars are available.

The first robotaxi services and tests in Europe are already showing how this could work. The company that develops the autonomous driving technology does not always own or operate the cars. The app where customers book their rides may come from another company too. This could leave an important role for taxi, rental, car-sharing and other fleet companies that already know how to manage vehicles.

What changes when the driver disappears?

A taxi company today manages both cars and drivers. Drivers need to be recruited, scheduled, paid and supported. With a driverless fleet, these tasks would eventually disappear, but most of the work related to the cars would stay.

Autonomous cars still need charging or refuelling, cleaning, tyres, inspections, repairs and insurance. They also need somewhere to park when they are not working. If there is an accident or a technical problem, someone needs to deal with it and get the car back on the road. Customer support is still needed too when there are problems with bookings or payments, or when someone leaves something in the car.

Some of these tasks may even become more important. Robotaxis are expensive, and every hour a car spends charging, being cleaned or waiting for maintenance is an hour when it cannot take passengers. Moove, for example, already provides fleet ownership, depots, charging, maintenance and 24/7 operations for autonomous vehicles.

Drivers also do more than drive: they notice damage, report faults, help passengers and respond when something goes wrong. Without a driver, these responsibilities need clear processes and dedicated support. Existing fleet experience is an advantage, but operators will need to adapt it to vehicles running without someone behind the wheel.

Who actually runs a robotaxi service?

There can be several companies behind one robotaxi ride. One provides autonomous driving technology, another may own the cars, and a local company can take care of the fleet. A platform such as Uber can then connect the cars with passengers. Sometimes one company may do several of these things, but this is not always the case.

Zagreb is one example. Pony.ai provides the autonomous driving system, Verne owns and operates the fleet, and Uber connects it with customers through its app. Autonomous rides became commercially available through Uber in August 2026 with a licensed operator onboard. On 10 September, Pony.ai and Verne also started fully driverless passenger test rides on a 22-kilometre route connecting parts of Zagreb with the airport. These are still limited test rides, not a fully driverless service across the whole city.

A similar model is being tested in Madrid. WeRide, Uber and AVOMO received permission in September 2026 to start testing and mapping under Spain’s ES-AV framework. The first 20 vehicles have an onboard specialist. WeRide provides the autonomous technology, Uber connects the cars with passengers, and AVOMO takes care of fleet operations. Commercial operations are planned for later in 2026, but the current permit is for testing and preparation rather than a general driverless taxi service.

In London, Wayve and Uber launched supervised autonomous passenger rides in September 2026. Customers can be matched with an autonomous vehicle through the Uber app, but there is still a licensed driver in the car.

Europe is preparing for larger robotaxi fleets

The plans for the next few years are already much bigger. Pony.ai and Uber announced in August 2026 that they plan to deploy more than 2,000 robotaxis in Europe, starting with Zagreb and later expanding to four more European cities.

Bolt and Lucid announced another partnership in September, with Bolt aiming to deploy at least 25,000 autonomous Lucid vehicles across European cities. This is a future target, not 25,000 cars already operating today.

How these fleets are owned and operated may also be different from one city to another. Local regulations, available vehicles, charging and depot infrastructure and partnerships will all have an effect on how each service is set up.

What could this mean for taxi companies?

Taxi companies already do many of the things that would still be needed with an autonomous fleet. They maintain cars, deal with insurance and incidents, provide customer support and make sure enough vehicles are available. Some also already have their own depots, charging infrastructure or maintenance partners.

One possible model can already be seen in Zagreb. A local company owns and operates the autonomous fleet, while a larger ride-hailing platform brings customers. In the future, an existing taxi company could potentially have a similar role. Instead of managing cars and drivers, it could manage autonomous cars that receive bookings through a larger platform.

This does not mean platforms such as Uber will always rely on other companies to own the cars. Uber’s 2026 agreement with Rivian includes up to $1.25 billion in milestone-dependent investment and says that Uber or its fleet partners may purchase the autonomous vehicles. Uber is also developing infrastructure for autonomous fleets. Different ownership models are therefore possible.

What about rental and car-sharing companies?

Autonomous vehicles could also become interesting for rental and car-sharing companies, although this is probably further in the future. These businesses already manage vehicles, maintenance, cleaning, charging, bookings, customer accounts and payments, so much of the practical work is similar.

Autonomous driving could also change how rental and sharing cars are moved. Today, an employee may need to drive a car from one location to another. In the future, the vehicle could potentially do this itself within an area where autonomous driving is approved. It could even become possible for a rental car to drive itself to a customer’s pickup point.

How do you measure a robotaxi fleet?

Without drivers, it becomes especially important to understand how much time each vehicle is actually working. Every hour spent charging, being cleaned, waiting for maintenance or simply sitting unused means less time carrying paying passengers.

Some useful KPIs could be available hours per vehicle, paid trips per vehicle each day, revenue per available vehicle/day, empty kilometres, charging and maintenance downtime, cost per trip and incidents that need human help. These are most popular robotaxi industry KPIs of what a company could measure during a pilot.

For example, if cars spend too much time waiting to charge, the charging setup may need to change. A lot of empty kilometres could mean the cars are not in the right places, while too much maintenance downtime could make a vehicle more expensive to operate.

How could local companies enter the market?

For either model, the business case depends on the partnership terms: who finances the vehicles, who carries the risk when demand is low, and how revenue and operating costs are shared. Operators should assess what remains after vehicle financing, insurance, charging, maintenance and partner fees. Depots, local service teams and reliable fleet availability could make an operator a valuable partner, but profitability still needs to be demonstrated.

One option for taxi fleets would be to work with an autonomous driving company and a larger ride-hailing platform. The local company could own or manage the cars and take care of the fleet, while customers book their rides through the larger platform.

Another option could be possible in cities where large ride-hailing platforms are not present or have limited coverage. A local company could work with an approved AV provider while using its own app, bookings, payments, customer support and fleet management.

In both cases, a company cannot simply buy autonomous cars and start offering rides. The technology needs to be approved, and local rules have to allow the service. These rules will also be different across Europe. Great Britain, for example, introduced its Automated Passenger Services permitting framework in 2026. Services similar to taxis or private hire vehicles need consent from the relevant local authority, while responsibility for the autonomous driving technology and the passenger service are treated separately.

What can fleet companies do now?

A company does not need to have autonomous cars today to start preparing. A good first step is understanding the current fleet in detail – what each vehicle costs, how many hours it is available, how much time it spends charging or being repaired, and when cars are used the most.

Charging, cleaning, maintenance and depot processes are also worth looking at because all of these will still be needed with autonomous vehicles. A company interested in an AV pilot can then choose a location and use case, look at the local permit process and start talking to autonomous driving technology providers.

The software will also need some preparation. Customers still need to book and pay for rides, while the company needs fleet information, customer support, reporting and connections between its own systems and the AV technology. Starting with a small pilot makes it possible to test all of this before trying to build a larger service.

Where could ATOM Mobility fit?

ATOM Mobility already provides many of the software functions used by taxi, ride-hailing, rental and sharing companies, including customer apps, bookings, payments, fleet management, maintenance tasks, analytics and APIs. Its ride-hailing software includes passenger and driver apps, dispatch, payments, fleet data, analytics and integrations. The ATOM Mobility API can also exchange data about vehicles, bookings, rides, payments, telemetry and maintenance tasks.

ATOM Mobility does not provide autonomous driving technology. This would come from a specialised AV company, and each project would have its own technical requirements and local approvals. For an AV pilot, ATOM Mobility could work with the fleet company and AV provider on the rider experience, integrations, fleet management and the data needed to measure how the pilot is working.

If you operate a taxi, rental or sharing fleet and want to explore an AV pilot,  ATOM Mobility can help plan the booking and fleet-management setup, integrations and metrics you would need to test together with an AV technology partner.

On October 8, 2026, we gathered on the panoramic top floor of the AC Hotel by Marriott in Riga for ATOM Connect Rental 2026.

Following the community-first approach of our past mobility summits, our goal for this edition was simple: avoid the noise, generic pitches, and sales booths of conventional trade shows. Instead, we brought together an intimate room of car rental operators, regional franchise managers, aggregators, and technology pioneers to have an open, unfiltered conversation about where vehicle rental is heading and what it actually takes to get there.

The day’s agenda reflected that intent. Rather than treating digitization as a buzzword, the sessions explored the operational realities of the industry: from global macroeconomic forces to the high-stakes friction points at the physical rental desk.

Here is a recap of the key discussions, industry data, and takeaways from the summit. 

The macro view: Three forces redefining vehicle rental (Frost & Sullivan)

Opening the morning keynotes, Frost & Sullivan set the strategic foundation by examining how vehicle rental is shifting from an asset-heavy model to a digitally enabled, flexible mobility ecosystem.

Their analysis highlighted three forces currently reshaping the industry:

1. Disruptive Technologies: App-based journeys, automated digital identity verification (AI KYC), and keyless access are quickly becoming baseline customer expectations.
2. Internal Cost & Margin Pressures: Repair costs, vehicle turnaround times, insurance inflation, and fragmented software are squeezing fleet utilization. Connected operations are now critical to protecting margins.
3. Transformative Megatrends: Electrification, changing travel patterns, and sustainability regulations are forcing operators to rethink fleet acquisition and pricing strategies.

The main takeaway from their presentation was clear:

"The winners will not simply manage more vehicles - they will manage the customer, fleet, and ecosystem more intelligently."

Looking ahead, Frost & Sullivan stressed that future value will come from experience-led rental models (low-touch, personalized journeys), commercial and LCV subscriptions, and ecosystem partnerships that maximize revenue per vehicle, customer, and trip.

A front-row reality check: Why car rental is the toughest service in travel (EconomyBookings)

Following lunch, Vitalijs Cvetkovs, Head of Commerce at EconomyBookings.com, delivered a refreshingly candid, data-backed session on why car rental remains the most friction-heavy customer journey in travel.

While booking a flight or hotel is relatively straightforward, car rental forces the customer to decode dozens of brands, ACRISS vehicle codes ("or similar"), deposit rules, insurance tiers, and fuel policies. Drawing on data from over a million annual bookings, Vitalijs pinpointed the four friction points where the industry loses customers:

• Choice Overload: Searches returning 700+ vehicle variations across 70+ brands overwhelm customers. Aggregators must prioritize smart, predictive ranking to surface the 10 most relevant cars rather than an endless list.
• The Broken Pricing Reflex: Unlike flights and hotels, car rental prices often drop in the final weeks before pickup. This trains travelers to book speculatively, monitor prices, cancel, and rebook - driving up cancellation rates across the board.
• The Credit Card Wall: EconomyBookings' audit revealed that less than 30% of European counter offers accept debit cards. With modern travelers increasingly relying on neobanks, turning customers away at the desk because they lack a credit card for the deposit creates major operational friction. Low-deposit and debit-friendly models are the inevitable direction of travel.
• The Counter as the "Emotional Summit": The counter is where booking expectations collide with operational realities. Their review data confirms an uncomfortable pattern: the lower the daily rental rate, the rougher the expected pickup experience.

Vitalijs concluded with a crucial commercial insight: In the era of AI and instant review summaries, poor customer experiences are instantly visible. Customer experience is the only sustainable retention driver, and aggregators and operators must work together to absorb friction rather than pass it on to the traveler.

Ground-level Intelligence: Spatial AI and next-gen insurance

Beyond macro trends and customer journeys, the technical sessions focused on practical tools operators can deploy today:

• Anadue (Urban Mobility at the Speed of Thought): Mike Manchip showcased their SuperFlexMap platform, demonstrating how spatial data drives branch profitability. By combining census statistics, commuting patterns, hotel density, and car ownership data, operators can perform gap analysis to identify affluent, car-free neighborhoods that lack rental desks - replacing intuition with empirical site selection.
• Datafolio (Unlocking Value in Add-On Insurance): Datafolio addressed the headache of unpredictable fleet coverage, showing how AI risk modeling (UBRA) and transparent, app-based micro-insurance can turn fleet insurance from an operational cost burden into an EBITDA generator.

Operationalizing the future: The INVERS fireside chat

Closing out the afternoon, ATOM Mobility hosted a fireside session with INVERS to discuss the nuts and bolts of deploying "digital rental-ready" fleets.

The session focused on practical execution. Using real-world examples, the discussion highlighted how telematics, CAN-bus hardware, and remote immobilizers enable traditional operators to launch virtual stations, service off-airport locations, and automate replacement vehicle workflows without increasing staff overhead.

The key takeaway: modernizing a fleet does not require an immediate, risky transition away from the counter. Hybrid operations allow established rental businesses to introduce 24/7 keyless handovers alongside their standard operations.

The Value of the Closed Room

As presentations concluded and conversations moved into the evening networking mixer overlooking Riga, the feedback mirrored what made previous ATOM Connect events special: the value is entirely in the room.  

By capping attendance and curating an audience of active industry operators, conversations went deeper than high-level presentations.  A sincere thank you to all our speakers, partners, and guests who joined us in Riga. The vehicle rental sector is evolving quickly, and we are proud to provide the space where the industry comes together to map out what comes next.  

Stay tuned for our next ATOM Connect events by following us on LinkedIn!