Mainstream press seems to have published almost weekly articles on autonomous vehicle trials. The articles typically talk enthusiastically about the exciting technology being tested and an optimistic future world where huge safety, congestion and economic benefits will be delivered by these revolutionary vehicles. Some production vehicles already have technologies that enable their drivers to ‘hand over’ control to their vehicle to a limited extent. So the journey to fully autonomous vehicles seems inevitable. The interesting question is therefore not if but when this will happen.
Predicting an accurate timescale for when fully autonomous vehicles are a regular occurrence is challenging. Our roads system is a complex combination of diverse physical infrastructure, rules and regulations, and (not always logical) road-user and pedestrian behaviour. And this takes place on a continuous basis across an extensive network, often in the harshest of weather conditions. Taking a successful autonomous vehicle trial into the real world and ‘productionising’ it will require inter-related areas to be addressed and there being stable capability relating to:
- Vehicles systems and technologies
- The regulation of the vehicles and their use
- The insurance of autonomous vehicles and clarity around where liability lies when things go wrong
- Compliance and enforcement of legal/regulatory obligations
- The demands/views of individuals and organisations of autonomous vehicles (both their use and interaction with other road users)
- How highways authorities and public authorities interface with and cater for autonomous vehicles.
Mass adoption of autonomous vehicles will only be possible when all these elements have been sufficiently addressed – “nothing is solved until everything is solved”. We have summarised some of this complexity in the diagram above (click on image to enlarge).
To help understand what is needed and when this might occur, PA Consulting developed an autonomous vehicle maturity framework. This framework is structured around the key capabilities and end-states that are required for fully autonomous vehicles to operate on our public roads effectively and with public confidence. The framework is technology/solution-agnostic – it doesn’t set out how the end-state will be achieved or what technology is necessary; rather it defines what capability is needed to achieve the end-state. To develop the framework, we drew upon the insight of PA experts in technology, software, systems engineering, business modelling and behavioural psychology. The framework was then validated in discussion with a range of senior industry stakeholders. On this basis the diagram below (click on image to enlarge) summarises the top-level capabilities we believe are key for fully autonomous vehicles to operate on our roads.
Using this framework, we sought to understand how close the UK is to having fully autonomous vehicles operating on our roads. To do this, we interviewed over 50 key stakeholders who had first hand involvement in autonomous vehicles in the UK. We were not looking for a statistically significant volume of answers. Rather we targeted individuals who we knew were working in the area of autonomous vehicles and therefore would have valid insight. Participants included senior individuals from:
- Automotive manufacturers
- The automotive supply chain
- Vehicle insurers
- Highways authorities (both local & national)
- Engineering organisations
- Police and other enforcement authoritiesLaw firms involved in autonomous vehicles
- National and Local Government (including those responsible for policy)
- Motoring organisations (representing motorists)
Each stakeholder was asked for their view on the current level of maturity for each of the capability building blocks in the framework. We also sought their views on how long they thought it would be before each of the capabilities would reach full maturity in the UK. The responses were collated, analysed and validated through a series of workshops, which included both industry and PA experts.
The first key finding was that stakeholders recognised that the framework provides an extremely useful mechanism for considering the complexities of autonomous vehicles and their use. A small number of enhancements were suggested to the framework, however there was consensus that such an approach was extremely useful. In particular it allowed individuals from a diverse range of organisations that historically have not had much interaction to consider autonomous vehicle roll-out using a common, technology-agnostic set of capabilities.
Our analysis showed that not all capabilities across the framework are equally well-developed/mature in the UK. For example, those capabilities around vehicle technologies were viewed as progressing well, however issues around how the police would deal with autonomous vehicles were seen as very immature (largely because they have not been considered). This is not unsurprising – in the inter-related system we would not expect all capabilities to develop at the same rate. The challenge here is that some elements of the system (for example around legal and regulatory) could have long lead-in times (for example in the development and the enactment of legislation). Further, as this is a system decisions in some areas will have impacts on others (i.e. there is feedback and dependencies). For example requirements around enforcement and insurance (which could be set out in legislation) could have an impact on technology decisions.
It is also worth recognising that some elements of the framework will be driven at an international level. For example economies of scale will mean that automotive manufacturers will look to develop vehicles that can be sold/used anywhere and not, for example, just for the UK market. Other elements will develop at a national level, for example legislation, the testing regime and how/if roads infrastructure is enhanced to facilitate the use of autonomous vehicles. In theory this national-level complexity would seem to make the adoption of autonomous vehicles more challenging, however it presents opportunities for good practice to be developed on the basis of different approaches, driven by a desire for individual countries to be the leaders in autonomous vehicle roll-out. We are currently running similar maturity assessment in other countries – early analysis confirms this.
One area that our analysis highlighted as particularly under-developed relates to user buy-in. Much of the press implies an almost insatiable desire of road users for autonomous vehicles. But this desire has yet to be tested in terms of benefits and, perhaps most importantly, costs – what will a journey cost? Is the service resilient? Will I enjoy ‘driving’?
Our analysis also highlighted that some areas, while relatively immature at present, could rapidly develop. For example, the capability for a member of the public to insure a fully autonomous vehicle is very immature at present – reflecting the fact that you cannot buy such a vehicle. However the financial services sector believes they can rapidly develop appropriate insurance products.
In terms of time to full maturity, our research suggested that it would be around 10 years before fully autonomous vehicles were regularly operating on our streets in the UK. Some of the stakeholders we spoke to were more positive about the timescale to full implementation – the most optimistic suggested three years. While the least optimistic suggested over 20 years. A key factor in these timescales was the possible impact of disruptors that is companies that have not typically operated in the automotive sector. While some saw the involvement of the likes of Google as accelerating the development processes, others highlighted that while they started to address some of the technology challenges the wider challenges still exist.
The path to having fully autonomous vehicle on our road is a complex one. The maturity framework provides a tool to help map out that path. And our recent assessment suggests that whilst we are making progress towards the goal, there is still significant work that needs to be done across a number of areas. Developing new technological capability is to be applauded. However this needs to be alongside and not at the expense of other capabilities that are key to having fully autonomous vehicles on our roads. ◆