Proof Over Vision: Mobility+AI in Reality Check

For years, the debate around automated driving was dominated by visions of robotaxis, software-defined vehicles and autonomous urban transport at scale. At the Mobility+AI Conference in Munich, that language was still present, but the tone had changed. The central question is no longer whether highly automated mobility will arrive, but under which conditions it can be deployed safely, approved by regulators, operated economically and accepted by society.

Why evidence now matters more than ambition

One of the clearest messages from the conference was that technological innovation on its own is not enough to carry the argument. The sector is moving beyond the phase of bold claims and entering a more demanding stage in which safety, performance and market readiness must be demonstrated in ways that are methodically robust, statistically defensible and regulator-ready. In that sense, safety is not merely a system feature but a claim that requires substantiation across multiple layers of development and operation.

That shift also changes how the industry thinks about approval and validation. As software-defined vehicles continue to evolve after start of production through updates and new AI-enabled functions, the old idea of type approval as a one-off milestone at the end of development loses credibility. If functionality changes continuously, safety cannot be tested only at isolated points in time. It has to be supported by a structured system that links data, model development, validation and field behaviour across the full lifecycle.

The conference made clear that this has strategic and economic implications as well. Companies that still associate market readiness primarily with additional test mileage or more isolated validation exercises are working with a logic that is becoming less effective. What matters increasingly is the ability to build a continuous evidence chain that connects data quality, model behaviour and real-world system performance. For manufacturers and suppliers alike, that means the real bottleneck is no longer only model capability, but the ability to make safety arguments internally coherent, externally defensible and regulatorily reusable.

Uber integrates robotaxis from Zoox

Robotaxis developed by Amazon subsidiary Zoox will soon be bookable via the Uber app. Services are set to launch in Las Vegas later this year, with Los Angeles following in 2027.

Benjamin Müller

Where automated mobility is most likely to scale first

Another striking theme at the Mobility+AI Conference was the greater precision around realistic application fields. Many speakers treated freely operating robotaxis in dense urban mixed traffic as a longer-term scenario rather than the most immediate route to scale.

Instead, the strongest near-term potential was seen in operating environments where boundaries are clearly defined, risk can be managed and the value proposition is concrete. That includes automated parking, logistics, fleet operations, shuttle services and other well-scoped infrastructure settings.

Automated Valet Parking, in particular, stood out as an example of how technological maturity, standardisation and regulatory alignment can begin to reinforce each other. It was discussed less as a showcase for innovation and more as evidence that automated mobility can move into operation when the use case, the proof structure and the approval logic are closely aligned. This marks an important change in emphasis. The most promising path into the market is not necessarily the one with the highest technical complexity, but the one with the strongest fit between benefit, safety case, liability regime and approval pathway.

That also helps explain why advanced Level 2 systems continue to attract major investment. Even manufacturers already working on Level 3 technologies are still pushing Level 2 forward because these systems are easier to scale and more readily compatible with current regulatory conditions. For business leaders, the implication is clear: capital allocation in this phase follows scalable deployment rather than technological maximalism. The quickest route into the market does not automatically run through the most ambitious autonomy stack.

Why regulation and trust are becoming strategic assets

The conference also underlined that automated driving is no longer shaped by any single law, standard or approval framework. It is governed by a broader regulatory ecosystem that includes type approval, international rulemaking, technical services, operational monitoring, liability, insurance, cybersecurity, software updates and the question of how field data feeds back into future safety arguments. Any company still treating compliance as a separate downstream workstream risks underestimating the true complexity of industrialisation.

This matters particularly in Europe, where the discussion often places greater emphasis on interoperable infrastructure, standardised validation methods, fleet-based applications and public transport use cases than on headline-grabbing commercial autonomy pilots. At first glance, that may appear less spectacular than the approaches taken elsewhere. Yet it may also prove more durable, because Europe’s strength could lie less in staging autonomy quickly and more in embedding technical systems within a resilient framework of regulation, public infrastructure and standardisation.

At the same time, the Mobility+AI Conference made clear that the next phase of automated mobility will not be decided by technical capability alone. Public trust is becoming just as important. Automated functions are still often seen by users as convenience features rather than safety gains, while tolerance for machine error remains lower than for human drivers. This makes acceptance a strategic category in its own right. Trust cannot be built through messaging alone. It depends on understandable safety architectures, clear accountability and reliable rules for failure cases. In the end, automated mobility will succeed or fail not only as a technology programme, but as a trust framework for real-world transport.

The Mobility+AI Conference 2026 sets the benchmark for what comes next

What lingers after the Mobility+AI Conference is the sense that the sector has matured. The benchmark for progress is not the most ambitious promise, but the quality of the evidence, the handling of regulatory complexity and the ability to translate innovation into clearly defined and commercially viable operating environments. That may sound less dramatic than the grand narratives that dominated previous years. For the industry, however, it is the more consequential message.

Only when evidence, regulation, liability and operational logic work together does a technological possibility become a scalable business model. That is the point at which autonomous mobility now stands.

Sponsored

Automotive Software Strategies Europe

19-20 May, 2026 SZ Tower, Munich