A shared car for the last mile, driven in turn by workers and residents

A conversation with Adriano Alessandrini, associate professor of Transport at the Department of Civil and Environmental Engineering of the University of Florence, scientific lead of the TUSS project and partner in the spin-off TUSS srl

Interview by Paolo Barbato with Adriano Alessandrini.

TUSS, The Ultimate Sharing Service, grew out of a simple idea: making car sharing and car pooling work together in a single service to reach the railway station. The first demonstrator ran between April and June 2024 in Sesto Fiorentino and the Calenzano industrial area, with nine electric golf carts approved for road use. The following year the service moved to a new setting, first between the university sites of Santa Marta and Morgagni, then, from July to September, along the route between Firenze Rifredi station, the neighbourhood and the Baker Hughes site. The trials were funded by the National Centre for Sustainable Mobility (MOST) with PNRR funds.

Behind TUSS is Prof. Adriano Alessandrini, who conceived and coordinated CityMobil2 between 2012 and 2016, the largest European initiative on fully automated road transport systems. In 2025 he co-founded the spin-off TUSS srl, with the aim of bringing the services it had tested to market.

On 18 September Alessandrini attended the launch of the territorial coalition of Mobility Managers taking shape in Florence. The member organisations read their employees’ mobility demand together, to find solutions that none of them could sustain on its own.

From demonstrators to the market

Paolo Barbato: Professor, TUSS was born at the University and became a company. How did that happen?

Adriano Alessandrini: The project started at the University of Florence, with a fairly large group of collaborators. We knew how to build the algorithms and had already developed them for simulations. Through a competitive call from the National Centre for Sustainable Mobility we obtained our first funding, bought the vehicles and turned those algorithms into a service people could use, first in Sesto Fiorentino and then for students of the University of Florence.

The results led the University to launch a pre-incubation process. The company was set up, took part in the Hubble acceleration programme and received its first investment from a US venture capitalist, with whom we developed the technologies in a production environment. We can no longer put the demonstrators’ golf carts on the road, because the municipalities that have to authorise them prefer such light vehicles not to travel alongside the rest of the traffic, and the university uses them for other trials, including automated driving.

The first commercial service, with conventional vehicles and customers paying the fare, is running in the area between Scandicci and Lastra a Signa, in the Metropolitan City of Florence.

Paolo Barbato: What results did the demonstrators deliver?

Adriano Alessandrini: One extraordinary result is the return rate. Convincing someone to use the service for the first time is not easy, but 80 out of 100 people who have tried it keep using it systematically. The other figure that reassured us is vehicle occupancy. The student service between Santa Marta, Morgagni and Rifredi station, which ran longer than the others and worked best, averaged 3.4 passengers per vehicle, while a private car usually carries 1 or 1.2 people. On the last mile one of our vehicles replaces three, and for the rest of the journey those people are on the train, so we take those cars off the road entirely.

Car sharing and car pooling in the same vehicle

Paolo Barbato: Why integrate car sharing and car pooling rather than offer them separately?

Adriano Alessandrini: On-demand services meet travellers’ needs very well, but they are expensive to run. Sharing services, on the other hand, work where demand is isotropic, with trips spread evenly in all directions, and therefore not in the suburbs, where they would be needed most. By combining two sharing services you get a third, which covers the last mile at a sustainable cost. The vehicle belongs to the service, as in car sharing, while one of the passengers drives it, as in car pooling.

Paolo Barbato: How does a trip work?

Adriano Alessandrini: People register as drivers or as passengers, and for each group the system chooses who drives, usually whoever is going furthest, so that they drop the others off along the way and are the last to get out. The vehicle is parked at the station. The group it has been assigned to gets off the train, takes the vehicle and, before reaching their destinations, picks up a resident who needs to go to the station. Once the workers have got out, the resident becomes the driver and takes the vehicle back to the station, so the vehicle never stops. Switching who is at the wheel is one of the most complex technologies we have developed, because we need to know where the vehicle is and check the safety conditions.

If a vehicle is left where nobody can reposition it, one of our operators steps in. They know in advance where it will be left, get there on a folding e-bike, load the bike on board and take the vehicle to where it will be needed next. We expect to have one full-time operator for every ten vehicles, and it is this ratio that makes the service financially sustainable as well.

The counter-flow resident

Paolo Barbato: What happens when only a few residents make the reverse trip?

Adriano Alessandrini: The vehicle has to be repositioned empty, and that cost pushes fares up too. The service for employees is an excellent start, but we need at least one resident travelling in counter-flow for every seven employees. Already in Sesto Fiorentino most requests came from workers, while those from residents were fewer. In the first commercial service a ride from Lastra a Signa station to the Scandicci industrial area costs 1.30 euros, and once more residents use the service the cost per ride will be able to come down.

The other costs are high too, starting with insurance. For one of our vehicles it costs 2,900 euros a year, because it covers passengers, driver, legal expenses and comprehensive cover for drivers who are not regular drivers of that vehicle. Even so, fares remain below the price of a public transport ticket, which costs 2 euros in Florence.

Who drives

Paolo Barbato: How willing are users to drive for others, often strangers?

Adriano Alessandrini: We measured it by interviewing 300 people. They split between those who prefer to drive and those who prefer to be driven, and the former are fewer than half. Of those who prefer to drive, around 70% also accept being a passenger, while those who prefer to be driven are more reluctant to become drivers.

These are statements, and statements often express wishes more than behaviour. In the demonstrators we saw that it depends a great deal on people’s attitude. Engineering students had no trouble switching from driver to passenger. Workers resisted a little more at first, but over time they get used to it and do it willingly. In the app, those who make themselves available to drive can also indicate that they would rather not, and in that case they are chosen as driver only if there is no alternative.

We are introducing mutual ratings between users who share a ride. Each vehicle also carries a black box that records lateral acceleration, harsh braking and acceleration, and minor impacts, and an algorithm derives a driver score from it. For now we only use it internally. We are considering sharing it confidentially with each driver, together with tips for improving, and when we have many drivers to choose from, those with a better score could be preferred as drivers.

Where it works

Paolo Barbato: What features does an area need for the service to work?

Adriano Alessandrini: You need a public transport hub between 1 and 5 kilometres from the destinations, a tramway or a well-served station, with a train every quarter of an hour. If a train comes every two hours, it is as if the station were not there. Around it you need businesses on one side and residential areas on the other, so that there are people getting off the train in the morning to go to work and people who live in the area and need to get to the station. Between Lastra a Signa and Scandicci, residents, businesses and public transport hubs are all within about ten kilometres, and that is the ideal condition for our current model. In Campi Bisenzio, on the other hand, the nearest station with many trains is Sesto Fiorentino, which is too far away, and the industrial and residential areas are separate, so it is hard to get the mix of demand we need.

We don’t want to go into Florence’s limited traffic zone (ZTL), because there even a shared car is one too many. In the centre people get around on foot, by tram or by bike, except those with mobility difficulties.

Paolo Barbato: In July Sergio Savaresi described autonomous driving to us as an extension of the rail network towards smaller, on-demand vehicles. You have been working on automation since the days of CityMobil2. Why did TUSS start with user-driven vehicles?

Adriano Alessandrini: There are many services of this kind in Europe, and almost all of them are built around a technological development. We did the opposite. We started from urban outskirts and peri-urban areas, where people who live or work there today have no alternative to the private car, and we designed a service that can be delivered with the technologies and the legal framework that already exist. In CityMobil2 the vehicles received their missions from a centralised management system, and TUSS adopts the same logic, but one of the users is at the wheel.

At the University we keep working on a technological leap that would make the service simpler. The University of Florence has patented a mechanical coupling device that forms a convoy of semi-automated vehicles led by a single driver, up to six vehicles if each is 3 metres long or a little more. With one operator repositioning six vehicles at a time, one resident for every thirty-five or forty workers would be enough, and areas like Campi Bisenzio would become suitable for a service of this kind. First, though, we needed a company able to run the service efficiently, in agreement with local businesses. The technology to make it work better, and elsewhere too, comes later.

The role of companies

Paolo Barbato: What role can local companies play?

Adriano Alessandrini: We discovered it with the first demonstrators. Companies make it possible to inform many people at once. A Mobility Manager, within the limits of privacy, can tell all employees about the launch of a new service, whereas reaching residents takes very expensive advertising campaigns or contacting people one by one. For us, companies have become the entry point into an area.

They are also the way to grow. Almost all sharing services have run up against the initial investment, because buying or long-term leasing many vehicles means spending four or five years with loans or lease payments not yet paid off.

With companies, they can be the ones to provide the vehicles to be shared, or to pay a contribution that covers the cost of a few vehicles. To stay in service sustainably, a vehicle must guarantee at least fifty individual trips a day. A company with two hundred interested employees might need ten vehicles, and even if it provided three or four, it would let us grow much faster than waiting until we could buy them ourselves. To assess a site you need the origin, destination and time of employees’ trips, the information underlying any dial-a-ride problem.

Paolo Barbato: What are the main obstacles?

Adriano Alessandrini: The main obstacle is a very rigid bureaucratic and organisational system. While we were negotiating with some companies over providing their vehicles, the taxation of company vehicles changed with the 2025 budget law (Law no. 207 of 30 December 2024, article 1, paragraph 48), and everyone stopped, because not even they knew what effect it would have. We work within a defined legal framework, but one that nobody had ever pushed this far, and it happens that one company’s legal expert and another’s give opposite answers on sharing company vehicles.

Image matters too, because some companies do not want their vehicles shared with others under any circumstances. For an engineer, used to thinking in black and white, this fragmented picture is the hardest part, but I believe it will resolve itself once the service becomes commonplace.

The rest is the condition of any startup: finding funding, and finding at a fair price the skills for activities far removed from the core work. In Italy, and in Europe in general, almost nobody invests significant sums in a startup from the outset, and many banks only grant loans to buy machinery, whereas a startup doing something new usually builds the machinery itself. We chose a sector that brings together public administrations, companies, transport and roads, one of the most regulated there is.

Habits

Paolo Barbato: How much do habits matter in the choice to give up the car?

Adriano Alessandrini: Those who manage transport are used to building the supply, but it is demand that decides, that is, each person’s individual choices. Living with a car or without one, or with one per family instead of three, is a very personal choice that takes a long time to mature. I myself, a transport professor with a home in the historic centre of Rome, only gave up my car after leaving it parked in the street for a long time, for fear of not finding a parking space again. Once I got rid of it, I realised how much of a problem owning it had been.

With the Municipality of Florence we conducted 15,000 interviews with citizens, and over 60% of those who travel by car, when asked how much they spend to do so, answer zero. That is why it is easier to shape the habits of younger people, who do not have a car yet. Companies can foster a different culture, that of getting to work without a private car, by offering TUSS, paying for public transport or organising a shuttle, and perhaps using the car park space for a nursery.

What it means for a Mobility Manager

Paolo Barbato: What can a Mobility Manager expect from a service like this?

Adriano Alessandrini: I am optimistic because our model holds up even when demand is not high. A traditional car sharing vehicle, to avoid making a loss, has to be rented 25-30% of the time, on average 6 or 7 hours out of 24. TUSS covers almost all its costs with the service for workers and residents in the morning and afternoon peaks, and during the other hours of the day and at weekends the vehicle can provide other services that bring in revenue. We can survive and grow even in an area heavily dependent on the private car, because we are integrated with the resources of companies and of the area. We can be one of the tools for reducing car ownership, but on our own we are not enough. We work well in certain areas, with technology we will be able to work in others too, and in the meantime we work here.