Automated vehicles are NOT the answer

Do you think automated vehicles are just going to come onto the market and solve our mobility problems? It’s much more complex, POLIS explains:  

Vehicle automation is possibly the most talked about development in the transport domain in recent decades. Headlines are not confined to the transport specialised press; they are widespread in the general media too. If you were to believe everything you read and heard about vehicle automation, you would be forgiven for thinking that driverless vehicles will be on our roads in the next 5-10 years and they will solve the main transport problems of congestion and safety. Is it so simple? POLIS does not believe so.

POLIS is concerned about the optimism bias in terms of what automated vehicles can deliver by way of benefits and when they will actually hit the roads. While automation may bring benefits, there is also the possibility that their widespread introduction in urban areas could lead to increased congestion, negative environmental impacts and negative health impacts, if walking and cycling are discouraged.

This was the backdrop for the discussions among POLIS members over the course of 2017, which culminated in a POLIS discussion paper setting out the perspective of its member city and regional transport authorities on the theme of road vehicle automation. Some 22 POLIS members contributed directly to this 12-page document, issued on 23 January 2018, which was endorsed by the full membership. In publishing this paper, POLIS hopes that research, technology development and policy-making will take into account some of the concerns, issues and questions raised.

The paper does not reflect a position but rather explores the potential impacts of driverless cars across a wide range of transportation domains and identifies some issues that city and regional transport authorities need to address and engage on as automated motoring advances. Given the uncertainties around when driverless cars will enter the market and in what form and the level of public acceptance, it is very difficult to predict the impacts. The paper therefore highlights potentially positive and negative outcomes across the domains that are relevant to city and regional transport authorities, particularly their policies of promoting sustainable mobility. In conducting such an exercise, POLIS members are better equipped to think about the policies and measures they could adopt to achieve positive outcomes and to avert or mitigate negative ones.

The impact of automated vehicles (AVs) on travel behaviour is one area that holds great unknowns: At one end of the spectrum where the car (private or shared ownership) remains an important mode of transport, there is growing consensus that there will be an increase in traffic and in km travelled. This would be to the detriment of public transport, walking and cycling. Indeed, such a scenario could lead to the end of high capacity public transport as we know it today.  At the other end of the spectrum, where vehicle automation leads to a reduction in private car use/ownership in favor of a positive mix of active travel (walking and cycling), high-capacity modes (train, tram and bus) and shared mobility, the outcomes could be positive. While this latter scenario appears attractive, it relies on wider factors such as willingness to give up the convenience and comfort of a private car and car-sharing economics – a fleet owner will want a fleet of vehicles operating all the time, even when demand is low.

The safe operation proposed by driverless vehicles on urban roads could in fact become an impediment to traffic efficiency

At a spatial level, self-driving vehicles are expected to free up road space and other land as a big part of on-street and off-street parking becomes redundant. However, any freed-up road space would need to be put to other uses, otherwise it would no doubt be taken up by vehicles. This creates an opportunity to redesign streets and to use that space more efficiently, such as different usages depending on time of day and type of demand. A downside to vehicle automation could be longer commuting trips by car since this time could be spent doing other things such as reading, working or sleeping, even in congested conditions. 

Vehicle automation holds potential to enhance transport accessibility among those population groups and areas that most need it, by reducing the cost of passenger transport provision in areas of low and dispersed demand (rural areas and suburbs) and for special transport services for the elderly and disabled. Current service provision may be fragmented and/or reliant on voluntary services. On the flip side of the social coin, it is not inconceivable that a future of market-led shared AV services could lead to different levels of service access depending on ability to pay: a premium subscriber could gain access to better and faster services, leading to increased social division and inequality.

Improving road safety is one of the key drivers for automating the driving task. Accidents due to driver distraction, a leading cause of accidents, should become a thing of the past in an automated future. Cities and regions see great benefit in vehicles programmed to comply with traffic rules, such as speed limits as they promise to be more cost-effective and efficient than current engineering measures (speed bumps, traffic calming, cameras, etc). However, since traffic rules and regulations can be specific to a given context and interpretation of those rules may differ from one city to another, a strong dialogue between industry and the public sector is required. With policies to promote active travel, cities and regions are particularly keen to know how automation can improve the safety of pedestrians and cyclists.

Given the complexity of the urban environment, with its short links, pedestrian crossings, on-street parking and range of road users – which is not the case for the ‘simpler’ motorway environment – there is concern that the safe operation of driverless vehicles on urban roads could in fact become an impediment to traffic efficiency. Will the sensors be so sensitive as to stop the vehicle regularly? What would be the impact on traffic flow of a mixed traffic environment involving vehicles of different levels of automation and active modes? The assumption today is that efficiency gains will only materialize once most (if not all) vehicles are highly automated – the long interim period could be painful. Further analysis and insights are needed.

With research and development mainly focusing on the vehicle environment, the requirements in terms of infrastructure are less clear. Ensuring a high level of service of the physical infrastructure, such as clear road markings and road signs, is difficult for a road authority today because of the cost involved and the fact that a utility company (telecoms operator, energy provider, water company) can dig up a road and may not leave it as they found it once their job is finished. 

Building the required digital infrastructure may also be expensive. There are expectations that many of the tools, such as digital maps and communication networks, will be developed and funded by the private sector, who will then need to find applications for them. Where the public sector is needed to make an investment, the cost and benefit will need to be carefully analyzed. For instance, road authorities are unlikely to invest in C-ITS purely for the purpose of enabling automation in the future. However, they are considering where C-ITS can play a role in supporting the traffic management task today, such as floating vehicle data and enhancing driver awareness of traffic rules. 

What’s C-ITS?
Cooperative ITS is a developing technology which will enable communication between vehicles and between vehicles and the infrastructure to deliver in-vehicle information (such as traffic light information or traffic information) and floating vehicle data to the traffic manager. Some vehicle manufacturers will start to make this communication technology a standard feature of new vehicles in the near future.  

A final domain of concern to city and regional transport authorities relates to legal aspects. While there are many discussions occurring at national and international levels, these are mainly focused on the vehicle and on creating the legal framework to allow automated vehicles to be piloted and even deployed on public roads. Such a discussion is missing at local/regional level, yet it is expected that new traffic rules will be needed for automated vehicles and new regulation may be required for the roll-out of automated car-sharing clubs. Specifically concerning the infrastructure, the liability issue needs to be clarified and EU rules may be needed regarding data sharing.

This is just a snapshot of some of the themes that need further attention as vehicle automation progresses. Building on these themes, POLIS members have identified a number of areas that they will give more attention to, namely:

• Build the right policy and planning framework to ensure they are prepared for and can steer the arrival of automated vehicles to ensure maximum benefit.
• Adopt a holistic framework, focusing on the services that automation can enable which lead to positive outcomes for city/regional transport goals and on most suitable road environment for such services
• Anticipate the potential changes in travel behaviour that driverless vehicles may induce and introduce measures to ensure this is in line with sustainable mobility goals.
• Be prepared for changes to the role and responsibility of the road/transport authority that will arise on the path towards the increasing connectivity and automation of vehicles.
• Ensure personal security and safety of those using automated vehicles, in terms of trust, intimidation and hacking.

The POLIS paper represents the start of a reflection process on how to move forward on the topic of road vehicle automation in a constructive way that harnesses the opportunity that automation offers in a way that supports the liveability and inclusiveness of Europe’s cities and regions.