Imagine hailing a bus like you would a taxi, but paying the same as a regular fare. Imagine one of these autonomous buses arriving at your door, sharing the ride with other passengers headed in your direction, and operating 24/24 without a driver: it's not a reality yet, but it could become one.
Maybe starting from Trento: a group of researchers presented a pilot project which includes ten 17-passenger autonomous shuttles (powered by hydroelectricity) for the city's historic center. If the model works and is scaled up, private cars would be replaced with shared autonomous buses. could cut CO₂ emissions by 92% per year.
Why Trento?
The choice of the city of Trento as a testing ground is not accidental. The Province of Trento has one of the highest motorization rates in Europe: 1.521 vehicles per 1.000 inhabitants in 2023, according to Eurostat data. More cars per capita than Germany, France, or Spain. This means more congestion, more emissions, and more urban space dedicated to parking instead of people.
The absurd thing is that many of these vehicles they remain still 95% of the time, while they could be replaced by a much more efficient shared system.
The project, presented to the Dolomite Conference on Climate Change and Sustainability, leverages existing infrastructure. Trento has already conducted autonomous driving experiments, has IoT-enabled traffic management, and participates in the European C-Roads corridor for intelligent cooperative systems. In practice, the city has roads that "talk" to vehicles, transmitting data on traffic lights, unexpected obstacles, and traffic flow in real time.
The plan includes the use of the Toyota e-Palette (See it in the cover photo): a compact electric shuttle (5m × 2m × 2,6m) capable of carrying up to 17 passengers. Designed for Level 2-4 automation, it integrates cooperative V2X (vehicle-to-everything) systems that allow vehicles to coordinate with each other and with the road infrastructure.
The first 10 autonomous buses would operate on 2-5 km circular routes in the historic center, and would be “recruitable” via an app, for 12-18 hours a day.
The mathematics of carbon
The project's numbers are impressive, but they follow a precise logic. Replace 100.000 private gasoline-powered cars with just 5.882 shared autonomous buses. would produce a net reduction of 92,02% in annual CO₂ emissions, equal to approximately 154.598 tons saved. I should point out right away that the calculation takes into account the entire cycle, ALSO including the emissions needed to produce the electricity that powers the buses. Okay? This way we avoid unnecessary questions.
The remaining 7,98% is the environmental footprint of the electric fleet itself: 189,84 grams of CO₂ per kilometer, resulting from 58 g/km of the vehicle plus 131,84 g/km linked to the carbon intensity of the Italian electricity grid in 2024 (412 grams of CO₂ per kWh). Because yes, let's say that electric cars are zero-emission., but Electricity has to be produced somewhere. Sure, hydroelectricity does its part, but there would be environmental benefits with any source.
After the replacement, then?
There comes a point when adding shuttles stops providing environmental benefits. Once all 100.000 private vehicles are replaced with shared shuttles, emissions savings reach a plateauFurther expanding the fleet would only add to residual emissions without actually eliminating new ones. At that point, the battlefront shifts: decarbonization of the electricity grid, route optimization, intelligent charging management, and increased vehicle occupancy.
The project includes three rollout phases:
Phase 1: pilots reserved for vulnerable users (elderly, disabled people, poorly connected extra-urban areas).
Follows the Phase 2: extension to school mobility, with apps that provide real-time tracking and personalized support for children with special needs.
Finally, Phase 3: full integration into the city network, with a hybrid pay-per-ride/subscription model and optional premium layers (priority reservations, express routes) introduced only after achieving rigorous standards of reliability, safety, and inclusion.
The project does not hide the short-term risks: the environmental footprint of batteries, public distrust of driverless buses, the digital divide for those who don't use smartphones, and the initial loss of jobs in the driver sector. But focus on long-term gains: drastic reduction in accidents caused by human error thanks to predictive algorithms, freeing up urban space currently occupied by parking lots, and above all new jobs in fleet maintenance, remote management and user support.
Autonomous buses: open software, public infrastructure
An interesting aspect of the plan concerns technological governance. While the hardware (the e-Palette shuttles) remains Toyota's property, Researchers propose making infrastructure software open-source: autonomous driving stack, data analytics layer, API framework. The goal is to avoid vendor lock-in, the mechanism by which a city remains locked into the technology choices of a single vendor.
Making the code public would allow global developers to contribute improvements (localized traffic algorithms, AI security protocols), would allow cities to adapt the system to their needs without relying on external updates, and would lower the barriers to entry for startups and emerging markets. Similar to what happened with Android in mobile phones or ROS (Robot Operating System) in robotics.
And where would the money to do this come from?
The estimated budget focuses primarily on fleet purchasing/leasing (~50% of total costs), digital infrastructure and sensors (~20%), charging stations (~10%), operations and maintenance (~17%), and research/impact assessment (~3%).
The funding would come from existing European programs: Driving Urban Transitions (44 million available for transnational projects), CCAM Partnership (1 billion budget in collaboration with Horizon Europe), Digital Europe Program for digital innovation projects.
These would be complemented by municipal and provincial co-financing, in-kind support from social associations (volunteers, access to user networks, awareness-raising activities), and corporate sponsorships from tech/telecom companies interested in testing solutions in real-world environments.
The project was also conceived to fuel a future IPCEI-CCAV (Important Project of Common European Interest on Connected and Autonomous Vehicles), an EU state aid instrument for cross-border projects when market mechanisms fail.
Autonomous buses: a model for export or a local experiment?
Researchers are under no illusions: success depends on many factors. Rigorous governance and integration are essential. compulsory With existing public transport to avoid rebound effects (more convenience = more trips = more congestion). We also need precise guarantees of fairness: territorial coverage requirements, fare caps, and access via SMS/phone calls for those who don't use apps. Transparent accident reporting is needed, with clear responsibilities. Furthermore, funded career transition plans for former drivers to take on fleet management, maintenance, and remote assistance roles. What do you think? Difficult, right?
Secondo McKinsey Center for Future Mobility, by 2035, autonomous mobility could generate $300-400 billion in annual revenues. But industry leaders themselves predict a 2-3 year lag (say 2038) before Level 4 urban robotaxis become a reality on a city scale.
Regulation remains the main barrier, development costs have increased by 30-100% since 2021, and value is concentrated in software (margins >15%) rather than hardware (~10%).
Self-driving buses: a test of trust
Global case studies teach us that technology alone is not enough. In ChinaApollo Go has demonstrated that tight vehicle-infrastructure-cloud integration works, but it relies on municipal permits and subsidized tariffs that are difficult to replicate in Europe without comparable regulatory frameworks. In Japan, national law enables the operations of autonomous bus operators (e.g. Eiheiji), but services remain limited to short rural routes on weekends and are reliant on subsidies. In the United Kingdom, CAVForth has put Level 4 buses into mixed traffic but with safety staff on board and union concerns keeping the unit economics fragile. In the United StatesFinally, Waymo has produced peer-reviewed reductions in crash rates over tens of millions of miles, but operates in defined ODDs and under constant regulatory and perceptual scrutiny.
Trento wants to avoid these bottlenecks by focusing on a different model: not private robotaxis only for those who can pay more, but public transport accessible to allNot an isolated experiment in a protected pedestrian zone, but a scalable system from the historic center to poorly served suburban valleys. No top-down technology, but co-design with vulnerable users, schools, social cooperatives, and retirement homes.
The project is still on paper, and the launch date hasn't been confirmed yet. But if it receives the necessary funding and authorizations, Trento could become Europe's first autonomous mobility project, designed to truly reduce the use of private cars, not to add a premium service to those who already get around well.
With 1.521 vehicles per 1.000 residents, the city has a clear problem to solve. Perhaps this is the right place to demonstrate that an alternative exists.
