Connected and Autonomous Vehicles: The future? Contents

Chapter 4: Further research

112.Whilst connected and autonomous vehicles (CAV) represent an opportunity in many sectors, the Government’s strategy to date has largely concentrated on the roads sector. In this Chapter we set out what research and development (R&D), commissioned by the Government and others, needs to be carried out over the short and medium term to ensure the benefits of the deployment of CAV are maximised and potential drawbacks minimised across the roads and other sectors, 63including agriculture and marine, where the potential benefits of CAV technology are considerable.

113.As far as road transport is concerned, there are two concurrent streams of development—one of connected vehicles and one of fully autonomous vehicles. It is likely that connected vehicles will be seen on UK roads in the short term (the European Commission strategy is working to a three year timeframe, see paragraphs 159–161) whereas fully autonomous vehicles will take much longer to appear.

114.In considering the necessary future research it is important that the Government plays to the UK’s strengths and that the research is not carried out, commissioned or funded by the Government where it is better carried out by industry or other stakeholders. The Government should not need to invest or take the lead in development of autonomous cars—this is best left to industry. However, the Government should continue to invest in the fundamental scientific research in robotics and information technology that underpins autonomous cars and other CAV, and also the social, human factors and network management problems that must be understood for deployment.

Testing facilities

115.We heard evidence about the three CAV road trials currently underway, part funded by the Government through Innovate UK (see Box 1), and the A2/M2 connected corridor trial run by the DfT in partnership with Highways England, Kent County Council and Transport for London. The Committee also visited the GATEway project in Greenwich and observed testing of the new autonomous pod vehicle.

116.Professor Nick Reed, Academy Director at the Transport Research Laboratory (TRL), and speaking on behalf of GATEway, stressed the importance of real-world testing environments for CAV, such as those used by the GATEway project.126 John McCarthy, Technical Director at Atkins’ Intelligent Mobility, and speaking on behalf of the Bristol driverless cars project, told us that there was an opportunity for the UK to “sell capability to the rest of the world” by bringing together a unified offering of test facilities and research projects.127

117.The Society of Motor Manufacturers and Traders (SMMT) said that more could be done to expand and improve the UK’s demonstration facilities and that the creation of a CAV testing facility would “help promote the UK as a world leader on [the CAV] agenda”.128 Furthermore, SmarterUK pointed out that current testing was focussed on urban centres and there was also a need for testing in suburban and rural areas.129

118.In July 2015 the Council for Science and Technology, in a letter about capturing value in the CAV industry, recommended to the then Prime Minister, the Rt Hon David Cameron, that the Government should work with business to create the world’s first testing facility in a real world environment in a busy UK town where CAV and their networks could be tested. The Council urged the Government to act quickly before the UK is overtaken by its global competitors.130

119.We note that the Government announced a prize fund for a town or city to develop a test site for testing of CAV in the Autumn Statement 2013.131 The Government then consulted on a proposal for a flagship test facility in June and July 2016 and in the Industrial Strategy Green Paper stated that it would publish its proposals in spring 2017.132

120.We are disappointed that the Government has delayed making a decision on a new flagship test facility. A delay is particularly damaging because CAV development is a fast moving area.

121.The Government must put together a comprehensive testing and research offer for CAV to attract manufacturers and academics to the UK immediately. This should include one or more large scale testing environments covering real world urban and rural environments.

Human factors—better understanding needed

122.Understanding how CAV will affect the behaviour of drivers, pedestrians, cyclists and other road-users will be important in developing both the technology that will underpin CAV and the policy for their deployment. The changes in human behaviour needed to interact with CAV could be significant. The evidence shows that there is still much work to be done in this area and that current knowledge is limited. The Government commissioned a scoping study to understand the main social and behavioural questions relating to CAV. This identified nearly 400 open questions and concluded that behavioural aspects of CAV have been under-researched.133

123.Professor Natasha Merat, from the Institute for Transport Studies at the University of Leeds, told us that, until recently, pedestrians’ understanding of CAV had been largely ignored because it is a very complicated topic—cultural and regional differences in pedestrian and cyclists’ behaviour are complex for human drivers to understand, let alone the sensors and cameras of a CAV.134 Professor Sarah Sharples, Associate Faculty Pro-Vice-Chancellor for Research & Knowledge Exchange and Professor of Human Factors at the University of Nottingham, pointed out that risks to CAV users’ safety may arise as a consequence of other road users adapting their behaviour in response to CAV being on the road.135 The TRL said that “pedestrians, cyclists, sensory impaired groups and other road users may also need to adapt their behaviour and expectations to accommodate the conduct of the various types of automated vehicle”.136 For example, it could be the case that pedestrians would become complacent and assume that CAV would avoid them, thereby crossing roads at any point.137

124.Level 3 autonomy (see Figure 1) may require CAV to hand back control of the vehicle to the driver when it is unable to deal with a certain situation. We received evidence suggesting that handing back in this way to a potentially unprepared driver could be very dangerous.138 Professor Neville Stanton explained:

“As vehicles become fully autonomous, even the most observant human driver’s attention will begin to wane. Their mind will wander … This is particularly true if they are engaging in other activities such as reading, answering emails, engaged in conversations with passengers, watching movies or surfing the internet.”139

125.Steve Gooding, Director of the RAC Foundation, expressed a concern that “at some point … the driver is not driving enough to be properly in control of the vehicle”. He said that he was concerned that a vehicle could hand back control to the driver who, for example, “might well be asleep at the time”.140 He went on to say that the risks inherent in a handover could be managed by skipping this level of automation altogether and requiring that an autonomous vehicle has to be capable of coping with any circumstances it encounters before being allowed on the road.141 Mr Wilson added that the handover is “a significant challenge … it is certainly a question we will need to answer before [the trials on the strategic road network] go live.”142

126.Professor Stanton told us that research had found that drivers of automated vehicles were generally not as effective in emergencies as drivers of manual vehicles. In simulated emergencies, up to a third of drivers of automated vehicles did not recover the situation, whereas almost all drivers of manual vehicles in the same situation were able to do so. In addition, research showed that drivers of automated vehicles took, on average, six times longer to respond to emergency braking of other vehicles compared to manual drivers.143

127.Professor Sharples said her research had found similar results.144 She said that it was important to understand the implications of increased autonomy on the capability of humans to maintain vigilance and attention in order to be able to respond to an emergency situation.145

128.Behavioural studies form part of the three CAV road trials outlined earlier (see Box 1), in particular the VENTURER trial in Bristol, Mr McCarthy told us was “looking to focus on the individuals, on understanding the behaviours and the relationship between people and autonomous vehicles, while not neglecting the need to understand the technology aspect of it”.146

129.Professor Reed told the Committee that as part of the GATEway project in Greenwich (see Box 1), his team were using a driving simulator to investigate the behaviour of drivers of manual vehicles in the presence of CAV.147 Furthermore, Professor Sharples advocated the use of simulation facilities, saying they were “critically important for both research and technology development. They have a very strong role in enabling researchers and companies to understand user behaviour and develop autonomous vehicles that are appropriate for use”.148 We had the opportunity to experience a driving simulator first hand during our visit to the GATEway project in Greenwich.

130.Professor Merat told us that it will be important to think about driver licencing for CAV and whether there will be a need for driver training, including for those drivers who already have licences for conventional vehicles.149 Furthermore, Professor Sharples said “we need to … maintain the understanding that people have an appropriate level of competence through a driving test. Even with fully automated vehicles we need to build in for the contingency that the driver will need to take control”.150 Professor Sharples also told us that there is a need to consider that a driving test may need to test an understanding of how a CAV will behave and not just control of a vehicle.151

131.CAV could have negative implications for drivers’ competence, making drivers complacent and overly reliant on technology. This is of particular concern in emergency situations, where a driver may react slowly to taking back control of a vehicle. It may be the case that for Level 3 vehicles the risks will be too great to tolerate. The risk of complacency also extends to other road-users who will interact with CAV, such as pedestrians and cyclists. Further research is necessary to understand these risks, including possible measures to address them.

132.We recommend that the Government should give priority to commissioning work to understand the main social and behavioural questions relating to CAV, in particular answering those questions identified by its own scoping study. This work should build on international research in this area. Furthermore Innovate UK and Government departments should ensure that studying behavioural aspects is an integral part of trials they fund, including access to simulation facilities if necessary.

Modelling of a mixed fleet

133.The impact of a mixed fleet on the UK road network, including the effect on congestion, needs to be understood. This understanding could be developed through the modelling of various scenarios, including different levels of penetration of CAV and variations in the degree of cautious or assertive behaviour by CAV, building on the modelling work carried out by Atkins for the Government in the January 2017 report Research on the Impacts of Connected and Autonomous Vehicles (CAV) on Traffic Flow. This would stand alongside the work done in real-life trials of autonomous vehicles (see Box 1).

134.The Atkins report concluded that the likely tipping point for the proportion of CAV to produce major traffic flow benefits may be between 50% and 75%.152 The Government told us that the research had shown that “CAV offer major potential to reduce delays, improve journey times and improve journey reliability on strategic and urban road networks”. However, the findings also highlighted that “these benefits are not a given”, and depended heavily “on the proportion of CAV in the fleet”.153 The report also concluded that initially, early models of CAV would act more cautiously and that the result could be a “potential decrease in effective capacity”.154

135.Mr Capes highlighted why this kind of modelling work is important, saying that “most local authorities … have very little understanding of how a mixed fleet could operate on the UK road network”.155 Dr Hermann Meyer, Chief Executive Officer of ERTICO-ITS Europe, indicated a need to think carefully about a mixed fleet transition period and to develop a clear strategy for it.156

136.We asked a number of our witnesses, including Mr Wilson, Mr Capes and Mr Hurwitz, about modelling, but they did not mention any specific work in this area in the UK or the work on micro-scale modelling carried out by Atkins on behalf of the Government.157 Mr Forbes told us that the Centre for Connected and Autonomous Vehicles (CCAV) had done some modelling, but it was still at a very early stage. He went on to say that CCAV are keen to develop that work in the future and to build the evidence base over time.158

137.We welcome the modelling work commissioned by the Government. This should be a starting point for further work on mixed fleet modelling to inform policy development. This work should help to counter the possible disadvantages and negative effects of managing a mixed fleet of autonomous and non-autonomous vehicles. The research on human factors must feed in to the modelling work so that as more is understood about human interactions with CAV, the modelling work can be refined to give more accurate results.

138.We were surprised that the work the Government has commissioned on micro-scale modelling was not more widely known by our witnesses. The Government should take steps to ensure future work in this area is more prominently planned and shared with stakeholders including LTAs.

Agriculture sector

139.Professor Blackmore told us that CAV could lead to changes in how farming is carried out in the UK, with the benefit of making the crop production system significantly more efficient.159 Professor Newman gave the example of a vehicle that can ‘see’ and apply herbicide to just the weeds and not to the crop.160 SmarterUK told us:, “Advances in this technology have the potential to make agriculture significantly more efficient.”161 Professor Blackmore also told us that CAV have the potential to make smaller farms more economical because agricultural CAV are likely to be much smaller than current farm machinery.162

140.In 2013 the Government launched a long term agri-tech strategy with initial funding of £160 million for R&D.163 Further funding has since been made available through the strategy. However, Professor Blackmore told us that he was concerned that the funding under the strategy is diminishing and that there will not be enough research funding available in the agriculture sector to realise the potential benefits of CAV for the agriculture sector.164 He told us that he would like to see agri-tech as a named area for research within the Industrial Strategy, “because the opportunity for British agriculture is so huge. If the funding were to slow down or stop then a lot of this innovation would slow down and we would not overcome the disruption [to agriculture]”.165

141.The potential for improved crop production and reduced adverse impacts on the environment by the use of CAV in agriculture is considerable. The Government should fund appropriate R&D once a business case is made which demonstrates the advantages which will accrue.

126 Q 34 (Prof Nick Reed)

127 Q 33 (John McCarthy)

128 Written evidence from the SMMT (AUV0058)

129 Written evidence from SmarterUK (AUV0089)

130 Council for Science and Technology, Letter to the Prime Minister ‘Capturing value in the Autonomous and Connected vehicles industry: an ambitious plan for the UK’, (23 July 2015): [accessed 5 January 2017]

131 HM Treasury, Autumn Statement 2013, Cm 8747, December 2013: [accessed 9 February 2017]

132 BEIS and Innovate UK, Driverless vehicle testing facilities: call for evidence (26 May 2016): [accessed 5 January 2017]

133 Supplementary written evidence from HM Government (AUV0095)

134 Supplementary written evidence from Prof Natasha Merat (AUV0092)

135 Written evidence from Prof Sarah Sharples and colleagues, University of Nottingham (AUV0049)

136 Written evidence from the TRL (AUV0039)

137 Q 6 (Ian Yarnold) and written evidence from Prof Sarah Sharples and colleagues, University of Nottingham (AUV0049)

138 Written evidence from Prof Sarah Sharples and colleagues, University of Nottingham (AUV0049)

139 Written evidence from Prof Neville Stanton, University of Southampton (AUV0029)

140 Q 48 (Steve Gooding)

141 Q 49 (Steve Gooding)

142 Q 49 (Mike Wilson)

143 Written evidence from Prof Neville Stanton, University of Southampton (AUV0029)

144 Q 57 (Prof Sarah Sharples)

145 Q 56 (Prof Sarah Sharples)

146 Q 32 (John McCarthy)

147 Q 32 (Prof Nick Reed)

148 Written evidence from Prof Sarah Sharples and colleagues, University of Nottingham (AUV0049)

149 Q 60 (Prof Natasha Merat)

150 Q 60 (Prof Sarah Sharples )

151 60 (Prof Sarah Sharples)

152 DfT, Research on the Impacts of Connected and Autonomous Vehicles (CAVs) on Traffic Flow: Summary Report (May 2016): [accessed 9 January 2017]

153 Supplementary written evidence from HM Government (AUV0095)

154 BBC, ‘Driverless cars ‘to increase congestion’ says government’, (accessed 6 January 2017): [accessed 9 January 2017]

155 Q 41 (Darren Capes)

156 Q 21 (Dr Hermann Meyer)

157 Q 53 (Mike Wilson) and Q 45 (Darren Capes)

158 Q 2 (Iain Forbes)

159 Written evidence from Prof Simon Blackmore, Harper Adams University (AUV0034)

160 Written evidence from Prof Paul Newman, University of Oxford (AUV0041)

161 Written evidence from SmarterUK (AUV0089)

162 19 (Prof Simon Blackmore)

163 HM Government, A UK Strategy for Agricultural Technologies (July 2013): [accessed 12 January 2017]

164 Written evidence from Prof Simon Blackmore, Harper Adams University (AUV0034)

165 Q 14 (Prof Simon Blackmore)

© Parliamentary copyright 2017