«We're testing ourselves to death in Europe»

There are various paths toward a safe, meaningful, and scalable mobility system of the future. Philipp Wetzel, Co-Managing Director of the Zug Alliance and Managing Director of Amag Corporate Development, Innovation & Venture Lab, noted at the start of the conference: «That’s why we also need diverse perspectives—whether from business, politics, or manufacturers. Because only through dialogue can we make progress.»

It is not just the powertrains but mobility as a whole that is changing; new players are entering the arena and transforming the value chains as well. Now, we need to determine how to build trust and acceptance for autonomous driving, especially among future users. There is also the question of what kind of mobility we actually want in the future. «In Switzerland, we have a strong public transit system, good infrastructure, and exciting framework conditions. That’s why we want to talk not only about technology, but also about the system and how everything works together,» Wetzel continued.

Not a mobility problem, but a spatial one

Carmen Walker Späh, President of the Cantonal Government and Head of the Department of Economic Affairs for the Canton of Zurich, echoed this sentiment at the Future Mobility Forum 2026: «Electric cars aren’t the future—they’re the present of mobility. Even autonomous vehicles aren’t the future yet—it’s the system in which they operate.» The mobility of the future begins where public space is shared, she noted, since all road users are ultimately competing for this scarce resource. «Cities therefore have a space problem, not a mobility problem,» said Walker Späh. «In the future, it must be possible to use public space intelligently. And this is where autonomous driving plays a very special role, as it can cover the first and last kilometers with an autonomous on-demand service that effectively complements public transportation.»

Europe needs to get moving

«Self-driving cars only work as part of an ecosystem—not simply because manufacturers put them on the road,» says Andreas Herrmann, a mobility expert at the University of St. Gallen, summing it up. Despite more than 40 different projects, Herrmann sees Europe as clearly lagging behind: «We have the tech giants in the U.S. and China’s five-year plans; both have virtually unlimited financial resources. To make progress with a European version as part of public transportation, we therefore need to move beyond pilot projects. »We’re testing ourselves to death in Europe.« The HSG professor assumes that there will be only a few companies worldwide that supply systems for autonomous driving. Now, he says, Europe must build its own player; otherwise, it may end up supplying only the hardware in the future. »And that apparently isn’t the most lucrative part of the whole thing,” Andreas Herrmann says meaningfully.

Scaling Up in the U.S. and China

Augustin Friedel, Associated Partner for SDV by MHP, put it even more clearly: «Level 3 autonomous driving is dead. BMW and Mercedes have pulled out again. Currently, only Waymo has Level 4 vehicles on the road.» Furthermore, he noted that the large-scale rollout of autonomous driving is taking place outside of Europe—in the U.S. and China. «There are 3,500 to 4,000 Waymo vehicles on the road in the U.S. alone—that’s a factor of 100 compared to what we have on the roads in Switzerland and neighboring countries in our individual pilot projects,» Friedel noted critically. In the U.S. and China, it’s estimated that autonomous vehicles will have a market share of 1.6 percent by 2035. These are still very small numbers, so it’s not too late for Europe, but we need to finally get moving now. «Because of the expensive sensor technology in robotaxis, they still cost around 100,000 francs today. To scale up production, these prices will need to come down to between 40,000 and 50,000 francs,’ the expert reveals.

Thinking Ahead Together with Public Transportation

This could also help lower the price of a single ride in a robotaxi and bring it closer to being economically viable. Nevertheless, it’s important to launch these services now in order to gain experience and attract customers. «Doing nothing is not an option,» explains Christian Willoch of Ruter, the public transportation provider from Norway. In the Oslo region, Ruter is focusing on an on-demand service with up to 15 autonomous vehicles. In the long term, the plan is to expand the fleet to 100 to 500 shared autonomous vehicles in public transit. «Mobility in our region will increase by 18 percent by 2050. We therefore need to achieve more passenger-kilometers per car, which will also help us combat traffic congestion,» says Christian Willoch.

But this whole concept doesn’t have to be developed outside the public transit system—it needs to be developed in conjunction with it. «Because from 2012 to 2018, public transit ridership in San Francisco dropped by 15 percent, and the main reason for that was ride-hailing services,» he explains. «So we don’t need to establish a service outside the public transit system, but rather integrate autonomous driving into the system. We need to rethink customer service and move toward sharing robotaxis—rather than having them drive around with just one passenger.»

Self-driving cars need space, too

Christian Heid, Vice President of Product, Service Design, and Operations at Benteler Mobility, also brought the Holon Urban—an autonomous minibus with a top speed of 60 km/h that can carry up to 15 passengers—to Zurich. The minibus with its big, round eyes is already in service in Jacksonville (U.S.) and Hamburg (Germany). However, the autonomous vehicle is only part of the solution—much more is needed. «Cleaning, charging, maintenance, as well as repairs and recalibration, are just as necessary for autonomous vehicles. With 25 Holon Urban vehicles, we need about 1,400 square meters for a depot; an additional 700 square meters are required if the fleet is expanded by 25 units,» explains Christian Heid. «Operating 25 vehicles also requires about 0.25 megawatts of electricity.» That’s roughly equivalent to the energy consumption of 50 to 80 modern single-family homes. In addition, real-time charging strategies should ideally be used so that weather conditions and the distance traveled the previous day are taken into account to ensure the most efficient charging possible.

The expert from Benteler Mobility also emphasized that, as a provider of such robotaxis, one must ensure every single level of safety—from vehicle safety to mobility safety. «When it comes to sharing, passenger safety also becomes a factor,» said Christian Heid. «Nine percent of people say they would never share a robotaxi. And a good third want to know who they would be sharing the ride with, which is likely to be difficult for data protection reasons alone.»

A Country of Pilot Projects

Autonomous driving is also being promoted in Switzerland, but politicians and federal agencies are currently focused primarily on «Verkehr ’45,» the Federal Council’s plan for expanding Switzerland’s transportation infrastructure. «It allocates nine billion for roads and 36 billion for rail,» explains Christa Hostettler, director of the Federal Office of Transportation. «Roads are barely mentioned in ’Verkehr »45,’ and that’s a problem. No one wants to get their fingers burned when it comes to road issues, but roads are exactly what we need.” Switzerland is a country of pilot projects and was an early adopter with the autonomous PostBus vehicles in Sion, but very little has happened in the last ten years.

«We see enormous potential for autonomous driving in shared mobility. But without a revision of the laws, no one will invest in this area,» Hostettler reveals, «since public transit isn’t allowed to turn a profit or even earn interest on its investments.» Rail alone cannot meet mobility needs, however; it is simply the backbone of the system, «but we need feeder services—and therefore also investments in roads—to make use of that backbone,» says the director of the Federal Office of Transportation.

Conflicts of Interest Identified Through Simulations

Here in Switzerland, officials are currently exploring what the mobility infrastructure in Lucerne might look like in the future. To this end, Salina Schmid, Lead of Innovation Scouting at Amag Lab, and Patrick Abegg, mobility coordinator for the Canton of Lucerne, are using the MATSim software and a digital twin of the Canton of Lucerne that includes its mobility infrastructure and key locations. «If the price for autonomous ride-pooling drops from roughly Uber’s level to that of public transit, demand will double,» says Schmid. At the same time, she notes, there will clearly be a need for more autonomous vehicles. Result: Revenue does not increase to the same extent. The simulations also revealed conflicting objectives between service quality, economic viability, and the transportation goals of such an autonomous ride-pooling system. So there is still much to be clarified. The ARiVE3 project, which aims to bring automated ride-pooling to Basel’s “Three-Country Corner,” is still in the feasibility and simulation study phase.

Still on the road with a safety driver

The «iamo» project has gone a step further: the Swiss Transit Lab, the cantons of Zurich and Aargau, and SBB are testing the use of self-driving vehicles in the Furttal valley. Project manager Pascal Kern reveals: «We have three robotaxis in operation, but they still have a safety driver on board. Hopefully, it’s a matter of weeks—not months—before we achieve fully autonomous driving. But I don’t want to talk about specific dates; I’d rather discuss phases. That’s something else I’ve learned from this pilot project.» Kern expects that the pilot will soon be expanded to include three nine-seat minibuses: «The technology is there, but »tech-ready« doesn’t yet mean it’s ready for rollout. We still need to learn a lot.» For example, autonomous driving turns a vehicle project into an IT project as well, one that must overcome additional hurdles due to regulatory frameworks and the entire data processing process. “An important lesson: We need larger pilot regions to make progress; even the entire Postauto AG is actually too small for a pilot project,” Pascal Kern states matter-of-factly.

Automate Logistics Rather Than Passenger Transport

Anna Chirkova of the Amag Innovation Lab then pointed out that the focus should not be limited to automated passenger transport; rather, freight transport in particular offers great potential for automation. In Asia, they are already scaling up operations in this area and aim to increase the fleet from the current 10,000 robo-delivery vans to one million by 2030. One of the largest users of these robo-vans is—unsurprisingly—the package delivery service DHL. For Amag and its logistics subsidiary Mobilog, automated delivery is a very exciting field, as there will also be a shortage of drivers in this country in the future. The number of drivers is projected to be 75,000 by 2050. One potential solution: delivering replacement parts via robovans.

«To do this, we used MATSim to create a digital twin of Mobilog and first fed it historical data on demand, routes, and orders,» explains Anna Chirkova. «Then we let it choose between using automated or conventional delivery vehicles and had it calculate various scenarios, looking at parameters such as empty runs, kilometers driven, and fleet size.» The assumption here is that, in the B2B sector for small and medium-sized shipments, it’s reasonable to expect the recipient to handle unloading if it can be done easily. «We sought to create added value for the recipient through fast delivery and for the logistics provider through the automated fleet, fewer vehicles, and fewer kilometers.»

For now, it's still altruism

For Amag CEO Helmut Ruhl, these investments in the various pilot projects are still altruistic at this point, «but we at Amag have an interest in moving forward and identifying a business model here.» And Philipp Wetzel, Managing Director of Amag Lab, summarized: «Looking at today’s discussions, we need to start thinking in terms of projects involving more than 100 vehicles. And we don’t have much time left, given the rapid developments in the U.S. and China. However, we could still become a model for Europe in the field of autonomous driving and thereby retain high-tech jobs in Switzerland.»