Twenty twenty(-five) Vision
Underpinning the Government’s thinking is its 2025 vision for CAM:
“By 2025, the UK will begin to see deployments of self-driving vehicles, improving ways in which people and goods are moved around the nation and creating an early commercial market for the technologies. This market will be enabled by a comprehensive regulatory, legislative and safety framework, served by a strong British supply chain and skills base, and used confidently by businesses and the public alike.”
The Government plans to realise this vision, which is set out in its Connected & Automated Mobility 2025 paper, through three pillars of activity:
Ensuring safety and security of CAM
Self-driving vehicle technologies and services will not be successful without certainty for innovators and investors, and confidence for the public that the technologies are safe, secure and work in the interests of society. This first pillar will be assured by “embedding safety and security by design”. This will involve introducing a comprehensive regulatory, legislative and safety assurance framework, further details of which are discussed below.
Government also intends to develop and implement new safety and cyber security assurance processes, noting that the technology underlying self-driving cars could make these vehicles susceptible to cyber-attack. As to the question of "How safe?", Government expects that "self-driving vehicles would be expected to achieve an equivalent level of safety to that of a competent and careful human driver."
Securing the industrial and economic benefits of CAM
Government’s objective to develop both the commercial market for CAM technology and a strong UK based supply chain will be supported by a further £60m of newly announced R&D funding (£40m of which is targeted at the deployment of CAM services and £20m towards supply chain capabilities). This investment, alongside Government’s other initiatives, including by way of its arms-length and industry bodies (for example the UK Automotive Council and Logistics UK) will seek to leverage private investment, drive innovation and assist businesses in addressing challenges through CAM technologies.
Delivering the societal benefits of CAM
Finally, stakeholder engagement will be key in order to meet the needs of end users and contribute to policy objectives. Government’s focus is to address public concerns and build understanding, trust and willingness to use self-driving technologies. This will involve conducting research on the public’s priorities and engaging with partners to ensure integration of CAM into wider networks, including roads, mass transit systems and wireless networks. In addition, Government has stated that it intends to promote CAM deployment in line with its “Future of Transport” principles (and environmental commitments).
A New Regulatory Framework
Setting the regulatory framework is a key next step in achieving the Government’s vision. The UK has already passed the Automated and Electric Vehicles Act 2018 (AEVA) but, despite the broad scope implied by its title, the main substantive feature of this legislation (as far as automated vehicles are concerned) was to create a direct right of action against insurers of automated vehicles. The scope of the regulatory framework now planned by the Government will be much more comprehensive, hence the requirement for further primary legislation.
Most of the Government’s proposals are based on the comprehensive joint report prepared by the Law Commission of England & Wales and the Scottish Law Commission. We have set out some of the key concepts below.
Levels of automation
The Government recognises that development and innovation in automated vehicles will be an incremental process, and is keen that regulation should not block this. Three main aspects of the proposed framework are designed to help with this:
First, it allows for the assessment of individual automated driving features rather than just at the level of the whole vehicle (although the whole vehicle will still need authorisation wherever it includes one or more self-driving features). The precise definition of an automated driving feature is not yet clear, but the Government sees it as “a function of the vehicle’s onboard technology, which would interact with the occupant(s) and may make certain demands of them” (this is slightly different from the Law Commission’s definition as “a system that can drive a vehicle in a particular set of circumstances or “operational design domain”).
Second, the Government proposes that a particular feature could be authorised subject to “deployment conditions”, which may limit how the feature is used. They give the example of a vehicle where there is insufficient evidence showing that it can operate safely in low levels of light – in that case an authorisation may be given but subject to a condition which prevents the vehicle legally driving itself at night-time.
Third, there is a recognition that in many cases automation will not jump straight to Level 5 (full automation) and regulation needs to cater for different levels of automation. There is therefore a distinction between features which require a user to respond in some way, most obviously to take back control of the vehicle – “Category 1 features” – and those which do not – “Category 2 features”.
Authorisation and the Authorised Self-Driving Entity (ASDE)
The proposals introduce a new role for the “Authorised Self-Driving Entity” or “ASDE”. This would be the entity that applies for authorisation of a vehicle that incorporates any automated driving features. In one of the main conceptual shifts compared to the regime for conventional vehicles, the ASDE will have an ongoing duty to ensure that the vehicle drives safely and legally through its lifetime.
For this reason the authorisation of the vehicle will require approval of the standing of the ASDE as well as the vehicle and its features. In particular the ASDE will need to meet requirements relating to financial standing, repute, and co-operation with government and other involved parties. Other duties of the ASDE will include ensuring cybersecurity, providing data to regulators and collecting, storing, and sharing data to process insurance claims (in compliance with data protection law). A consultation process is expected to provide further detail prior to implementing ASDE duties by way of secondary legislation.
The authorisation process will build on existing work developing the ‘monitoring test’, which seeks to determine whether a vehicle can safely and legally drive itself without the need for monitoring by an individual. A vehicle must: (i) comply with relevant road traffic rules; (ii) avoid collisions which a competent and careful driver could avoid; (iii) treat other road users with reasonable consideration; (iv) avoid putting itself in a position where it would be the cause of a collision; and (v) recognise when it is operating outside of its operational design domain.
User-in-Charge (UiC) v No User-in-Charge (NUiC)
The ASDE will therefore have overall responsibility for ensuring that the dynamic driving of the vehicle is safe whilst an automated driving feature is operating. But the proposed framework would also include two additional actors carrying a degree of responsibility:
The User-in-Charge (UiC): in the case of Category 1 features this would be the person in the vehicle who, when prompted by a “transition demand”, has to take control and effectively become the driver, thereby taking on full responsibility for complying with road traffic rules. A UiC would therefore need to be licensed to drive. Even if an automated driving feature has control of the vehicle the UiC would have responsibility for ensuring compliance with traffic rules that do not relate to the dynamic driving of the vehicle (such as ensuring the vehicle is roadworthy, seat belts are on etc.).
The No User-in-Charge (NUiC) Operator: in the case of Category 2 features there would be no UiC, and therefore the Law Commissions have proposed the licensed NUiC operator as an alternative actor to assume responsibility for compliance with non-dynamic driving rules. The NUiC operator would act remotely, and therefore could not be someone inside the vehicle. Furthermore the NUiC operator could be the same as the ASDE but does not need to be: for instance a provider or public passenger services (buses etc.) could be the operator, as could someone who wishes to lease a fleet of vehicles to customers; where a vehicle with a Category 2 feature is privately owned it would be necessary, at least in the early stages of the regulatory framework, for the owner (or registered keeper) to engage a licensed NUiC operator’s services to oversee the vehicle.
The precise scope of the NUiC operator’s responsibilities remains to be defined: the Government stresses that the ASDE (not the NUiC operator) would remain responsible for the safety of the vehicle’s dynamic driving, but in practice the distinction between dynamic and non-dynamic may blur. The Government also makes clear that the vehicle’s passengers “can be expected to bear no responsibility for any element of the vehicle”; this seems slightly different from the Law Commissions’ approach of allowing for flexibility in scoping the NUiC operator’s responsibilities, for instance so that registered keepers take some responsibility for maintenance in certain cases.
The “in-use” regulation scheme will enable ongoing safety monitoring of self-driving vehicles throughout their life cycle. Government expects that data and evidence as to overall safety of self-driving vehicles collected through this regulation will inform safety standards. For instance, where a specific incident occurs involving a self-driving vehicle, or where a significant negative trend in traffic infractions is identified, a new body (the Road Safety Investigation Branch (RSIB)) will carry out investigations, identify learnings and emit independent safety recommendations. The in-use regulation scheme and incident investigation mechanisms are expected to act as a feedback loop for safety, driving “continuous learning” in the operation of these new technologies.
Data collected is expected to facilitate the following assessments:
- Whether the authorisation remains valid;
- Frequency of any traffic infractions (the meaning of which will be defined in legislation and/or guidance);
- Whether sanctions are appropriate against the ASDE (or, where appropriate, the operator); and
- Whether there has been a failure to disclose, or a misleading disclosure of, safety-relevant information, which will be a criminal offence by officers of the ASDE (or, where appropriate, the operator).
Although Government intends to work towards a ‘no-blame culture’, appropriate sanctions will apply on ASDEs where there is evidence of a lack of due diligence that led to an incident. These sanctions will scale from informal warnings through to compliance orders, civil penalties and suspension of authorisations.
Looking onwards, Government has mapped out the ambitious programme of activity to help deliver driverless vehicles by 2025 as follows:
- The “Connected and Automated Vehicles: Process for Assuring Safety and Security” (CAVPASS) programme will guide the development of the safety framework. This will be developed alongside the safety, cyber security and behavioural research programme set to take place over the next two to three years;
- The majority of primary legislation needed to clarify responsibilities and support a safety framework is now likely to come in a Transport Bill in the 2023-24 Parliamentary session. Presumably this delay also means that the consultations needed for secondary legislation and guidance will be pushed a year later, into 2024-2025; and
- Starting in 2023, commercial deployment pilots, mass transit feasibility studies and supply chain competition supporting the commercialisation of key connected & automated technologies, services and capabilities will ensure commercial services models are ready for investment by 2025.
Authored by Oliver Wilson and Matthieu Bouharmont.