What the London Assembly’s autonomous vehicle report means for the UK’s CAM ecosystem 

The London Assembly’s latest report on autonomous passenger vehicles raises important questions about safety, accountability and public confidence. Operators are already engaging with many of these challenges, but London is setting a high benchmark. For the UK’s connected and automated mobility sector, the opportunity is to translate that ambition into practical, collaborative approaches to continuous assurance and real-world integration. 

The prospect of autonomous passenger vehicles operating on London’s streets is moving closer to reality. But as the London Assembly Transport Committee’s recent report, Who’s driving the future of autonomous transport in London?, makes clear, successful deployment will depend on much more than the capabilities of the vehicles themselves. 

The Committee raises concerns around safety, congestion, accessibility and the potential impact on the transport workforce. It also questions whether the current regulatory arrangements are sufficient to ensure autonomous passenger services deliver benefits for Londoners. 

For the wider connected and automated mobility (CAM) ecosystem, these findings are not a suggestion that operators are starting from scratch. Waymo, for example, set out its own vision for London in September 2026, addressing safety, accessibility and how its service could complement the capital’s transport network. The Assembly’s report challenges the sector to build on such commitments and show how they can be evidenced in London’s operating environment. 

From demonstrating safety to continuously assuring it 

One of the report’s most significant challenges concerns the standard against which autonomous vehicles should be assessed. 

The national benchmark of a careful and competent human driver provides a basis for safety assessment. However, the Committee questions whether this is sufficiently ambitious for a city working towards its Vision Zero objective of eliminating deaths and serious injuries on London’s transport network by 2041. 

This distinction matters. 

Demonstrating that an automated vehicle can perform at least as safely as a competent human driver is not necessarily the same as demonstrating that its deployment will actively contribute to reducing road danger. 

It also raises a practical question for operators, policymakers and regulators to answer together: how can the longer-term ambition of Vision Zero be translated into measurable, achievable safety expectations for autonomous passenger services today? 

Zenzic’s work on verification, validation and assurance (VVA) reflects the importance of considering safety throughout the operational lifecycle of automated vehicles, rather than treating validation as a single hurdle to overcome before deployment. 

As systems encounter new environments, interactions and edge cases, operational evidence must inform ongoing assessment and improvement. 

Turning real-world data into meaningful learning 

The Assembly recommends that operators report crashes and near misses, including information about the vehicle’s decision-making processes, within 24 hours. 

This could provide an important foundation for continuous safety assurance. Operators already invest in safety processes and operational learning, but meeting London’s expectations consistently will require agreement on how evidence is defined, interpreted and used. 

London’s streets are exceptionally complex. Interactions between pedestrians, cyclists, buses, delivery vehicles and other road users create situations that are difficult to anticipate and interpret. 

Identifying what constitutes a near miss is not always straightforward. Understanding why an automated vehicle responded in a particular way, and what can be learned from that response, presents another challenge. 

This brings the importance of explainable AI and interpretable data standards into sharper focus. 

The CAM ecosystem needs approaches that allow operational evidence to be understood, assessed and, where appropriate, shared. Learning should contribute to improving safety across the sector, rather than remaining confined to individual proprietary systems. 

Developing common definitions, reporting approaches and mechanisms for sharing relevant insights could help make continuous assurance more effective. 

Designing services around people 

Safety, however, extends beyond how a vehicle navigates the road. 

The report also highlights challenges associated with removing the human driver, particularly when passengers require assistance or an unexpected situation occurs. 

How should an autonomous passenger vehicle respond to a medical emergency? What happens when a passenger feels unsafe? How can services accommodate people with different accessibility requirements? 

These are not simply questions to address once the technology is ready. 

Operators are already considering these questions as they plan services. London’s higher expectations reinforce the need to demonstrate how accessibility, safeguarding and interaction with emergency services are embedded in vehicle design and operational procedures, rather than treated as additions after deployment. 

Engaging disabled people and other users in service design will be essential to ensuring autonomous mobility delivers meaningful benefits, rather than creating new barriers to transport. 

A wider test for the UK CAM ecosystem 

London offers a demanding environment in which to demonstrate the potential of autonomous passenger services. 

But the lessons extend well beyond the capital. 

The Assembly’s report does not mean the sector has overlooked safety or public needs. Rather, it sets a demanding benchmark for showing how existing operator commitments translate into verifiable outcomes, continuous assurance, transparent operational evidence and services designed around Londoners. 

Meeting that benchmark will require collaboration among operators, regulators, transport authorities, disabled people and the wider CAM ecosystem. Together, they can develop proportionate measures, common reporting approaches and practical ways to assess progress towards Vision Zero. 

This is an opportunity to make autonomous passenger vehicles part of the roadmap to reducing road danger, while recognising that the contribution of these services needs to be measured and demonstrated in operation. 

The next step is a shared one: working with operators to turn London’s ambitions into clear operational expectations and evidence that autonomous passenger services can help move the capital towards Vision Zero.