Social Intelligence Report

Public attitudes to automated vehicles An update to Robotics and Autonomous Systems: what the public thinks, reviewing information on the views of the public on automated vehicles. December 2014 Highly automated vehicle: A car which is capable of operating on the road without human intervention, but is fitted with full driving controls, and in which a driver must be able and ready to assume control. (Definition based on Department for Transport, 2014) Fully automated vehicle: A car which is capable of operating on the road network without human intervention, and in which a driver need not be able and ready to assume control. (Definition based on Department for Transport, 2014) Autonomous vehicle: Vehicle installed with autonomous technology which has the capability to drive the motor vehicle without the active control or monitoring of a human operator. (Definition based on the Statutes of Nevada, 2013) Driverless car/Self-driving car: Terms used in mainstream media to denote fully automated and autonomous vehicles. Flocking technology: Systems in which individual vehicles are autonomous enough to operate independently, yet can function collectively as a group and synchronise through communication in order to minimise the information processing and actuation energy required for collision avoidance. (Definition based on H.G. Tanner, 2004; and B. Crowther, 2002)

Executive Summary Despite a relative lack of perceived urgency behind these developments, many see benefits to society including freedom of movement for the elderly and disabled, increased comfort and convenience for commuters, and improvements to congestion through the use of flocking technology. There are concerns around increased unemployment and decreased human contact as a result of automation, along with fears about cyber security and privacy around the currently involved organisations and governments. There are questions about how the technology will manage the complexities of real driving situations, including coping with unexpected obstructions, adverse weather and failed infrastructure. More information is demanded about how automated vehicles will coexist with current cars and predict human behaviour, and who is responsible for malfunctions that result in accidents and damage to people and property. Ethical issues are also raised, such as how people will accept the tragic consequences of an accident if there is no human to be held accountable. Negative perceptions of the technology are often associated with a desire to know more about it. Public introduction of automated vehicles might therefore ease public concern. Informing about the pressures on transport infrastructure might also help to put the benefits of the technology into context.

Introduction Automated vehicles use a range of technologies to augment or replace human driving. Technologies such as steering assistance and collision-avoidance systems and even fully autonomous vehicles could have far reaching impact on safety, productivity and employment. Sciencewise produced a report on public attitudes to Robotics and Autonomous Systems (Sciencewise 2013), after this was identified as one of the Eight Great Technologies (Royal Society 2012). This report focuses on the use of robotics in travel in automated or ‘driverless’ vehicles. Automated vehicles use a range of

technologies to augment or replace human driving. Technologies such as steering assistance and collision avoidance systems and even fully autonomous vehicles could have far reaching impact on safety, productivity and employment. The technology is close to public use, with the UK government allocating £9m to increase the prize fund for driverless car testbeds, enabling trials in Bristol, London, Milton Keynes and Coventry from 2015 (Autumn Statement 2014). However, there is an important ongoing role for public dialogue in the continued development and improvement of these policies as the technology and associated issue develops.

Public views

Public views on potential benefits

The public views summarised in this report have been extracted from the sources detailed in the appendix.

1. Providing freedom of movement for the elderly and disabled. Driverless cars could benefit specific groups; including the elderly, disabled and those too young or too old to learn to drive.

The quantitative data on public views on automated vehicles originate from The Ipsos MORI Loyalty Automotive Survey (2014), and the qualitative analysis of over 2,800 public comments published online in response to mainstream media coverage on the topic. In addition, the views expressed in the trade magazine TransportXtra have been taken into account. With the exception of the Ipsos MORI survey mentioned above, there is no data on public opinion on this topic available from the Eurobarometer or other Ipsos MORI studies. At the time of writing, the UK Government’s Department for Transport was in the process of analysing the feedback received during a public consultation on the regulatory framework for testing highly automated cars. This feedback could not be made available for the purposes of this report. General awareness: A brand new technology in the test phase Recent media coverage has focused on the fact that driverless cars will be tested on UK public roads from January 2015 (BBC, The Guardian, The Daily Mail, all July-August 2014), with the aid of £10m of government funding to be granted to Greenwich, Bristol and a joint project between Milton Keynes and Coventry (e.g. BBC, July 2014). Both highly automated and fully automated vehicles are mentioned as being reviewed for trial. Google’s tests of their ‘driverless’ vehicles in California have also received press coverage here in the UK (e.g. The Telegraph, August 2014). Some readers’ comments indicate that they are unaware of the level of automation already in use in current vehicles, and driverless cars are generally received as a completely new technology. Although there are a small number of people who are convinced the reported developments are just a “PR -exercise” or an excuse to get legislation amended for the purpose of testing military vehicles (comments on The Guardian, July and August 2014), the majority believe highly automated cars will soon be seen on streets in the UK to be used by (a subgroup of) the public. There are mixed ideas as to how long it will take for the technology to be ready for use. General public attitude: Lack of importance The Ipsos MORI Loyalty Automotive Survey (2014) shows that the majority of respondents do not consider the development of automated vehicles to be a priority. Only 18% of those surveyed said they thought driverless technology was an important avenue for car manufacturers to pursue, with 41% deeming it unimportant. This could be linked to the finding that improving urban mobility and easing congestion are not seen as “life or death” issues (Ipsos MORI : Hearing and Being Heard, 2013).

2. Comfort and convenience. Automated vehicles are believed to “take the stress out of travel” (comment on the BBC, May 2014). Increased comfort and a more productive commute are also mentioned, as well as the opportunity to have a drink and drive home safely. 3. Increased safety. Fatigue, drink driving and general human failure would no longer play a part in accidents. Many believe that drivers on the roads in the UK currently pose a large risk, and would feel safer in the hands of machines. 4. Easing congestion. The possibility for flocking or convoy technology may mean automated vehicles would be more efficient than traditional cars. More cars could fit on less road space, as they can drive at smaller distances from each other. The communication between vehicles could also mean that fast lanes could operate without a speed limit, and that traffic jams would be avoided. Parking problems would be solved as the cars could either continue to pick up other people, or drive back home alone. 5. Decreased cost. Increased efficiency could also mean lower costs for passengers, and the lack of drivers would reduce running costs for freight firms. 6. Environmental benefits. Flocking technology may lead to greater efficiency, thus decreasing environmental impact, whereas car sharing would mean less cars would be needed. “It really is hard to see the point of driverless cars.” (comment on The Guardian, July 2014) “Just because we can, doesn’t [sic] always mean we should.” (comment on The Daily Mail, August 2014) Despite the generally reserved attitude towards automated vehicles, many readers indicate that they can imagine certain advantages to the technology. The most commonly cited benefits are: Freedom of movement - “I am unable to drive for health reasons, making transport very expensive and inconvenient for me. This would give me my freedom and stop me having to rely on friends and family when public transport isn’t available.” (Comment on The Daily Mail, August 2014) Flocking - “Robotic cars pollute much less. Firstly, they are programmed to make less [sic] mistakes which cost a lot of energy, and secondly they can also be programmed to drive in convoy down freeways/highways sheltering each other in wind slips - the amount of energy saved could be huge. Thirdly they can have a ‘hive’ like intelligence which would prevent traffic gridlocks.” (Comment on The Guardian, May 2014)

“I have not met people who believe the self-driving car is upon us but many of these people I have met believe that one day it will arrive.” (comment on The Telegraph, September 2014) The general lack of perceived urgency is reflected in readers’ comments on news reports, and there is a sense that it is an example of tech for tech’s sake.

Public Attitudes to automated vehicles

December 2014

Concerns

Unanswered questions

There are also a number of recurring concerns expressed by readers. These can be summarised within the following themes:

In addition to the perceived lack of need for driverless cars, the cautious attitude towards them appears to be fuelled by a lack of clarity on how the technology would work in practice. The main uncertainties expressed can be divided into four types of questions:

1. The system is vulnerable to abuse. Terrorists may find ways to bring transport to a halt, or create unmanned car bombs. Pranksters may abuse the fact that the automated cars will be programmed to stop and constantly jump out in front of them. Hackers could play havoc with the software. Insurers and manufacturers may be flooded by fraudulent claims. 2. The degree of control big business and governments could gain over people’s lives. Google, a company that many people perceive as having invaded their privacy already, would know where drivers have been. Possibly in collaboration with the government, “they” could flood passengers with advertising whilst in the vehicle, drive them to where they are likely to make a purchase or even directly to the police station if they have committed a transgression. 3. Safety. Many readers compare automated driving technology to devices they are familiar with and invariably fail, e.g. computers and satnavs. They feel machines aren’t up to tasks where human lives are at stake. If the car requires a human to take over the controls in an emergency, the driver may be too out of practice to safely do so. 4. Unwelcome changes to lifestyle. The convenience of driverless cars could further decrease people’s use of their bodies, leading to health issues. The human contact between driver and passenger would disappear, and unemployment would rise, leading to a greater divide between the rich (who benefit from lower running costs) and the jobless. Those who enjoy driving would regret to see that pleasure taken away from them. 5. Waste of existing resources. The comfort and convenience of automated vehicles could mean the end of public transport, and a waste of the investment in for example HS2. Personal investment in driving lessons is also mentioned as potentially going to waste. 6. Environmental concerns. The comfort and convenience of driverless cars may make people willing to travel longer distances by car, increasing urban sprawl and rendering energy efficient public transport obsolete. In addition, if a license is no longer required, more people will have access to their own vehicles. Empty vehicles driving around would further add to the environmental impact. Hacking - “Nicking a few photos may amuse the odd pervert, but how much more satisfying would it be to send 500,000 London commuters to Scunthorpe?” (Comment, The Telegraph, Sept 2014) Tracking - “These cars provide another platform to track us... It’s no coincidence that Google are pushing this hard - they’re a marketing company who thrive on exploiting private [data] for profit. I’d rather retain control of my life.” (comment, BBC, May 2014)

Public Attitudes to automated vehicles

1. How will the technology manage the complex realities of driving? How will it match up to human information processing when it comes to distinguishing between a pigeon that will fly off and a “football that might be followed by a child?” (Comment The Telegraph, Aug 2014). How will it negotiate unexpected obstructions, and how will the sensors cope with adverse weather? How will it navigate considering satnavs are often wrong, and what will happen when a signal fails? 2. How will the new technology coexist with current cars? Computers are bad at predicting human behaviour, and won’t be able to make eye contact with drivers to aid communication. Human drivers may get frustrated with the rigid behaviour of automated cars. Flocking technology may mean that automated vehicles will be hard to overtake. Reduced reaction time of automated vehicles may cause pile-ups when human drivers respond more slowly. 3. Legal issues: Who’s responsible when it goes wrong? Will the programmer be liable? How will insurance work? Will you need a license to “drive” a highly automated vehicle? Who will pay for fines or take the points for misdemeanours? 4. Ethical issues: What will happen when an unmanned vehicle kills someone? Can people accept error with tragic consequences when a machine is making the decisions? There is a sense many might struggle if there isn’t another human being to be held accountable. “Too many variables for my liking.” (comment on The Daily Mail, Aug 2014) The UK’s role The suggestion that driverless technology affords the UK the opportunity to “lead the world in driverless technology” (The Guardian, July 2014) is met with ridicule by some readers, and there is a sense that the country is lagging behind. The £10m of government research funding is considered by many to be “a drop in the ocean”, indicating a lack of commitment to the technology (comments , The Guardian, July 2014).

Analysis of views Interest in automated vehicles appears to be low-to-moderate. This may reflect societal feelings towards the two broader areas which are brought together by the technology: transport and autonomous robotics. As a science topic in general, transport ranks mid-level when it comes to public interest and perceived benefit (OST/Wellcome Trust, 2000). In addition, autonomous robots are met with more reserve than other novel technologies, and are the only area of scientific research which was seen as beneficial by less than half of the population in the Public Attitudes to Science survey of 2008.

December 2014

However, with a large volume of news reports focusing on impending tests on UK roads, the majority of the public appear to realise that driverless cars may one day become a reality. In other words, the technology of automated vehicles has reached a “tipping point” where it is moving out of the “domain of science fiction” and into the real world (. S. Le Vine and J. Polak, 2014). The volume of online comments on the topic in 2014 compared to previous years may reflect a growing interest and concern. Despite a relative lack of perceived urgency behind these developments, many see potential advantages of automated vehicles. Imagined benefits to society include freedom of movement for the elderly and disabled, increased comfort and convenience for commuters, and improvements to congestion through the use of flocking technology. On the other hand, losing the need for drivers could lead to a rise in unemployment and a greater economic divide. Human contact could further decrease as a result of the technology, and loss of physical exercise could lead to health problems. Cyber security is another major worry, with many fearing the system would be too vulnerable to terrorism and abuse. There is also significant concern for privacy and a fear that corporations and governments may work together to gain an unacceptable level of control over citizens It appears that there are currently too many questions left unanswered about how the technology would be implemented for the majority of the public to be ready to embrace it. Legal issues, ethical questions and practical uncertainties need to be addressed before a majority might be willing to back the introduction of automation on our roads. Discussions Of great importance to many people is the issue of road safety. Opinions on this are strongly divided. Advocates of automation believe that by eliminating human error, automated vehicles pave the way to a completely accident-free society. On the other hand, opponents of the technology feel that driving requires complex human information processing and communication skills which no amount of programming can match. This group is reluctant to allow machines to fulfil a role where human lives are at stake. The debate on the competencies of man versus machine takes up a significant part of the online public discussion on driverless technology. Another area of discussion centres on environmental issues. Some believe automation will lead to increased fuel efficiency and reduced emissions, especially through the application of flocking technology and car sharing. Conversely, the comfort and convenience of driverless cars could fuel demand, increasing the burden on both infrastructure and environment.

MORI ’s finding that amongst those who are not interested in cars, the support for driverless technology is significantly higher at 25% (2014). This indicates that those who enjoy driving are less likely to be in favour of the introduction of automated vehicles. A second factor which may play a role is technological literacy. There appears to be a group of highly enthusiastic, technologically literate individuals who are ready for automated transport. The possibilities of automated vehicles seem to speak strongly to their imaginations; conjuring a detailed, almost utopian vision of the future. This corresponds to the OST and Wellcome Trust’s finding that those classified as “Technophiles” were more likely than any other group to be interested in new and faster methods of transport (2000). These tech-loving enthusiasts envisage that driverless cars will function as an advanced form of public transport. Shared pods will pick people up from their homes and join convoys when on larger roads, thus increasing efficiency and comfort and solving logistical problems. Some of these advocates feel, however, that the current automated technology has missed a trick by failing to make use of renewable energy sources. Trends The public views expressed online in 2014 do not seem to differ from those posted in 2010 in response to the news of Google’s first US trials. However, a far greater number of people have been taking part in these voluntary online discussions in recent months, which may indicate that awareness of and interest in the technology is on the rise. This is ikely to be a result of the news that tests will be taking place on UK roads in 2015, bringing the issue closer to home. This driverless car looks to stop most of that carnage, by eliminating human error. Let’s face it, human error is far more prevelant [sic] than people like to admit.” (comment on The Daily Mail, Aug 2014) “There are road signs which just say... “THINK!” No machine ever will. Novel situations? These machines are like the Victorian “Automatons.” (comment on The Telegraph, Aug 2014) “Ever stop to consider that manual driving is proven to kill millions of people already? “Self-driving cars will be one of the greatest changes in these modern times, world changing and excellent. It will be the biggest step forward to real civilisation.” (comment, The Guardian, May 2014)

Influencing factors The prospect of automated vehicles is met with a wide range of responses. One of the factors which may play a role in determining whether or not an individual will welcome the technology, could be the level of enjoyment they derive from driving. Many people comment on the fact that they simply like to drive, and would regret to see that pleasure taken away from them (e.g. The Guardian, 31 July, and comments). This corresponds to Ipsos

Public Attitudes to automated vehicles

December 2014

Implications

Gap Analysis

A large number of readers are asking questions about how automated vehicles would work in practice. This may suggest that addressing the confusions surrounding the technology could increase positive attitudes towards it.

There are several areas of research which could be explored to aid policy makers in implementing policy relating to automated vehicles. As this is a technology which is close to public use, tracking public attitudes to the technology as it becomes more widespread would provide important evidence for policymakers

Whereas testing on UK roads in the near future will involve highly automated rather than fully automated vehicles, the public is largely worried about the implications of completely autonomous cars. Clarifying a time-scale may therefore help ease public concern. Considering that the perceived need for automated vehicles is relatively low, the public may need to be encouraged to reflect on the consequences of future transport demands. A recommendation from Sciencewise’s report Hearing and Being Heard (2013) suggests that “participants will need to be encouraged to see the potential wider implications of very heavy traffic and more road accidents, e.g. loss of economic productivity, more burden on health system, or they may find it difficult to engage with this topic”. Findings from the Eurobarometer survey of 2008 suggest that video material and television broadcast may be the most effective way to introduce automated technology. “If technology is to do with fuel and cars, then put the information on something like the television program Top Gear.” (Eurobarometer 2008) (On an educational film about smart cars): “It showed very realistically how it all happens in a car. It is not so obvious when somebody just describes it in a brochure. In this film you see the benefits of those smart cars and it actually creates interest.” (Participant comments from Eurobarometer survey 2008)

Public Attitudes to automated vehicles

Stronger evidence - This report has highlighted concerns and opinions of the public, taken from a single quantitative study of opinions, and from comments taken from online articles. As there is little systematically collected evidence on public awareness and attitudes to automated vehicles, more studies on this would strengthen the evidence available to policy makers. Public acceptability - This study has identified the perceived benefits and risks of automated vehicles, but this is not necessarily a good indicator of whether they would be willing to engage with the technology. The Public Attitudes to Science report 2014 (Ipsos mori 2014) found that there was a big difference in public opinion when respondents were asked about use of automated systems in transport in general, compared to use of automated buses to transport people. It’s clear therefore that the attitudes to automated systems is highly dependent on their context – where they are driving and what they are carrying. Further study on the specific situations in which the public would or wouldn’t find automated vehicles acceptable and whether they would be willing to drive alongside automated or driverless cars would add to understanding of public attitudes. Changing landscape - The evidence presented here was collected at a time when automated vehicles is merely theoretical. As everyday use of automated vehicles is likely to become a part of daily life over the coming years, it would be valuable to track public attitudes to the technology as it becomes more of a reality, to see how this affects positive or negative perceptions of the technology.

December 2014

Sources of data • Only 18 per cent of Britons believe driverless cars to be an important development for the car industry to focus on, Ipsos Mori poll. 30 Jul 2014 • Automated Cars: A smooth ride ahead? Dr Scott Le Vine and Professor John Polak for the Independent Transport Commission, Feb 2014 • Expert opinion paper exploring the consequences of automated vehicles for Britian • NRS: Chapter 482A: Autonomous Vehicles, 2013 Statutes of Nevada, 2013 • Review of the legislative and regulatory framework for testing driverless cars, The Department for Transport, 2014 • Rule-based guidance for flight vehicle flocking, Bill Crowther, School of Engineering, Sept 2002 • Flocking with Obstacle Avoidance in Switching Networks of Interconnected Vehicles, Herbert G. Tanner, 2004 • Public Attitudes to Science 2008, By People Science & Policy Ltd/TNS Mar 2008 • Science and the Public: A Review of Science communication and Public Attitudes to Science in Britain, Research by OST and Wellcome Trust, Oct 2010 • Hearing and Being Heard, Ipsos Mori, Aug 2013 • Qualitative Study on the Image of Science and the Research Policy of the European Union, Eurobarometer, Oct 2008 • Google’s driverless cars ‘can’t use 99pc of roads, The Telegraph, 1 Sept 2014 • Google forced to add steering wheel to driverless cars, The Telegraph, 22 August 2014 • Google’s driverless cars ‘will be allowed to speed’, The Telegraph, 19 Aug 2014 • UK to allow driverless cars on public roads in January, BBC, 30 Jul 2014

Acknowledgements This report, and the others in the series, has been produced by Sciencewise. The author is Louis StuppleHarris, British Science Association and he can be reached at [email protected] Sciencewise would like to thank: • Amitti CanagaRetna, Department of Health • Helen Kuhlman, Technology Strategy Board

• Our Work Here is Done: Visions of a Robot Economy, NESTA, 2014 • Look, no hands! I’m in Google’s driverless car (and it could be finding its way to YOUR street) The Daily Mail 30 Aug 2014 • The driverless sleeper car is a dream in Technicolor, the Observer, 9 October 2014 • Driverless cars get green light for testing on public roads in UK, The Guardian, 30 Jul 2014 • Google’s driverless car: no steering wheel, two seats, 25mph, The Guardian, 28 May 2014

• Mike De Silva, Department of Health

• Driverless cars will ruin the thrill of driving, The Guardian, 31 Jul 2014

• Andrew Acland, Sciencewise Dialogue and Engagement Specialist

• Google is to start building its own self-driving cars, BBC World Service Radio, 28 May 2014

for taking the time to review draft versions of this report and for all their useful comments and insights. This report and others are available online at http://www. sciencewise-erc.org.uk/cms/what-the-public-say-3

Public Attitudes to automated vehicles

• Google brings driverless cars to the roads, The Telegraph, 12 May 2012 • Are driverless cars the future for motoring?, confused.com, 17 May 2012 • Google gets Nevada driving licence for self-drive car, BBC News, 8 May 2012

December 2014