What Type of Self-Driving Vehicle Do Citizens Imagine? Results of a Co-Creation Dialogue Across Five European Countries

Abstract

It is believed that the deployment of autonomous vehicles in Europe has the potential to achieve safer, more sustainable and more equitable mobility. This study provides new insight into the hopes and expectations of citizens across five European countries for autonomous transport systems via a qualitative co-creation activity. A diverse and inclusive sample of 147 citizens was invited to generate their own ideas about how self-driving transport could be usefully deployed. Participants were asked to describe, in detail, what the vehicle would be like, what kind of mobility service it would provide, and who the target users would be. Structured qualitative analysis of the answers identified 337 distinct scenarios for autonomous vehicles across seven archetypes. In this paper, we describe a number of features of autonomous services that citizens expect and analyze the differences between demographic groups. We found that citizens across all five countries share a vision of autonomous mobility as electric- or hydrogen-powered, with a much greater use of shared mobility than is currently the case. This understanding of the expectations of a diverse group of citizens offers valuable insights for industry and policymaker actors to implement regarding future technology and transport investment and policy and service development.

1. Introduction

With transport contributing around 5% to the Gross Domestic Product (GDP) and employing more than 10 million people, this sector is crucial for Europe’s economy [1]. Transport is also responsible for a quarter of European greenhouse gas emissions, making it critical to the goal of a 90% reduction needed to achieve the commitment of Europe to becoming climate-neutral by 2050 [2].

Multiple strategies have been adopted to ensure the competitiveness of the European transport industry at the same time that it is reshaping towards more sustainable, safe, and efficient mobility [3,4]. In this context, autonomous vehicles (AVs) have emerged as a new trend that, it is argued, can improve quality of life, increase the safety of roads, and provide more sustainable and inclusive mobility services [5].

Technology is evolving rapidly, and AVs are starting to transition from demonstration pilots to commercial deployments. This complex socio-technological transformation of the transport system must be attentive to questions regarding both social and technical systems to secure equitable public benefits [6,7]. In this regard, expert stakeholders have been involved in key projects funded by the European Commission to design future transport services enabled by automatization and vehicle functionalities through workshops aimed at reaching consensus among participants [8,9,10]. In terms of public engagement, a wide number of studies in recent years have explored the awareness, attitude, concerns, willingness, and expectations of citizens toward this type of vehicle, as well as the most desirable characteristics of driverless mobility. Studies include diverse methods (i.e., surveys, interviews, citizen dialogues, focus groups) applied to different types of samples in different continents [11,12,13,14]. There are also a number of studies that synthesize insights from user preferences and public acceptance of AVs [15,16,17] and factors that influence travel mode choice and travel behavior [18,19]. These studies typically take the form of presenting options or choices to individuals, describing scenarios generated by the researchers, and frequently reaching different conclusions about how individuals will choose to travel when self-driving vehicles become available.

Beyond these studies of individual preferences and responses to expert-generated scenarios, there is a gap in understanding of what diverse and inclusive groups of citizens imagine and expect from mobility services provided by autonomous vehicles when the scenarios themselves are generated by individuals. Co-creation is an increasingly widely used tool in transport development and research, as evidenced by the interest of European actors like Civitas [20] and their increasing prevalence in transport research, including on the topic of autonomous vehicles [21,22,23,24]. Pappers et al. [25] defined co-creation in transport research as a form of public participation involving creative methods and innovation, with high levels of citizen empowerment or control. As operationalized here, co-creation can provide insights into how individuals envisage the future, using open questions and prompts to allow research participants to lead the discussion.

The study reported here is part of the Horizon project Move2CCAM [26], which uses multiple methods of inquiry to understand and model the impacts of AVs. This paper reports on data collected from a diverse and inclusive sample of 147 citizens from five European countries through the deployment of a co-creation activity [27] inspired partly by [28], which demonstrated its potential to be replicable. It reports on the types of autonomous vehicles and services that citizens imagine could meet their transport needs in the short and medium term. It investigates the expectations of citizens within a range of demographic criteria and across five geographically distinct locations. In line with socio-technological transition framing, we propose that an in-depth understanding of the hopes and expectations of citizens should be the basis of the action taken by industry transport operators and policymakers to develop and deploy automated mobility solutions that meet genuine public needs so that they are widely accepted, and therefore, deliver on the potential benefits of autonomous vehicles.

2. Materials and Methods

The Move2CCAM project invites citizens across Europe to participate in the co-creation of a shared understanding of the role and impact of AVs in future transport systems. This concept was operationalized through the recruitment of a diverse cohort of citizens and organizations who took part in multiple co-creative activities over a period of 2 years. As co-creation methods are less standardized than other qualitative approaches [28], we provide a detailed account of our procedures for reference.

This paper reports on the findings of the first co-creation activity of the Move2CCAM project, which aimed to understand citizens’ hopes and ideas for how AVs could be deployed via a semi-structured online discussion. This format was chosen because it allows for the formulation of novel ideas while providing significant flexibility to pursue themes that emerge during the course of discussions. A tailored research protocol was defined for the creation of use cases by the Move2CCAM partners, which comprised a social research company with experience in co-creation processes (Think Insights), the research entity in charge of the analysis of the insights compiled (CARTIF), and the technical coordinator of the project (UCL). These partners collaborated to identify a sampling strategy for participation and design the co-creation activity and the analytical approach. A detailed manual for the co-creation event with instructions and recommendations to guide the moderators was used to support collaborative discussions among participants and consistency between the different locations.

2.1. Sample Characteristics

Participants were recruited from five European countries with different features, which were Germany (DE), Spain (SP), France (FR), Cyprus (CY), and the United Kingdom (UK). Stratified sampling was used to recruit public participants and achieve a sample that was inclusive of people with a range of characteristics relevant to mobility needs, in particular those known to be under-represented in existing research on autonomous transport. The recruitment criteria were as follows:

• Age: To address different factors that contribute to shaping each generation’s travel behavior, a range of participants was recruited, with 25% of participants aged 18 to 34, 25% over 65, and the remainder of adults between 35 and 64.
• Gender: An even split of men and women was recruited, noting that existing evidence shows variation in attitudes towards autonomous vehicles.
• Health Conditions: Greater accessibility for those with conditions affecting mobility has been claimed as one of the potential benefits of autonomous vehicles; this study, therefore, included 20% of participants reporting a disability or long-term health condition that limited their mobility, alongside any carers.
• Place of Residence: The mobility needs of urban and rural residents are quite different, and, to date, most autonomous services have been trialed in urban areas. Participants were, therefore, recruited living in rural, suburban, and urban settings.
• Driving Capacity: Increased mobility for those who are unable to drive is another stated benefit of autonomous vehicles; therefore, this study included at least 10% of participants who do not have a license or are prevented from driving by a health condition.
• Income: Information was recorded about income level (relative to national averages) for analysis purposes but not as a recruitment criterion.

Table 1. Demographic features of participants involved in the co-creation activity.

Features	Categories	# Participants
Age	18–34	32
	35–64	68
	65+	31
	Unknown	16
Gender	Male	65
	Female	66
	Unknown	16
Health conditions	Yes, additional needs needed	20
	Yes, carer for someone with additional needs	9
	No additional needs	82
	Unknown	36
Place of residence 1	City center	17
	City periphery	36
	Town	28
	Village < 2000 inhabitants	17
	Unknown	9
Driving capacity	Have a license and can drive	105
	Have a license but cannot drive	11
	No license	15
	Unknown	16
Income 2	High	43
	Medium	37
	Low	33
	Unknown	34
Total		147

¹ Participants from Germany were classified into different subgroups: urban areas, towns, and villages. ² Participants from Cyprus did not provide income data. #: Number of.

provides the demographic features of 147 participants from the study.

Table 2. Income ranges for citizens involved in the co-creation activity.

	DE, FR (EUR)	SP (EUR)	UK (EUR)
Low	<24,999	<27,999	<25,999
Medium	25,000–49,999	28,000–41,999	26,000–49,999
High	>50,000	>42,000	>50,000

reflects the income ranges in each participant’s country.

2.2. Co-Creation Activity

Participation was via an online engagement platform called Recollective, which allowed for asynchronous participation and a mix of individual and group activities (e.g., during the reflection phase). All activities were delivered in the local language spoken by the majority, and participants were supported by a member of the research team who provided technical support and moderated the online discussion, encouraging the completion of responses and seeking clarification where responses were ambiguous. Data were collected over the course of five days during March 2023 for a total of one hour and a half of engagement per participant.

The co-creation activity involved the following three phases:

• Setting the scene aimed to establish a common understanding of the basic concepts under discussion and to introduce the co-creation process. The activity started by asking participants about their lived experiences of transport by exploring the mode that they used for daily journeys, the challenges faced, the type of people most likely to suffer from transport constraints in the area where they live, and the improvements needed. Then, since there was no requirement for participants to have any pre-existing interest in or knowledge of AVs, the concept of autonomous vehicles was introduced with a short video on how an autonomous car operates without a human and a brief overview of different vehicle types under development. Finally, to support the objective of hearing from people with a broad range of experiences and characteristics, we introduced ten proto-persons that presented different mobility needs, desires, travel behavior, and socio-economic features.
• Idea generation: the core of the activity required participants to define potential mobility services that could be provided by autonomous vehicles in the area where they lived. Participants were able to generate multiple ideas and detailed descriptions of vehicle design and operation, as well as the kind of mobility service that should be provided. Participants were given a series of open-text questions to respond to as prompts, for example, what would this/these vehicle (s) be? What kind of mobility service (s) will it/they provide? What will these services allow you to do that you cannot do now? What kind of a person might need or use this service?
• Reflection: the final phase asked participants to review and comment on the ideas generated by their peers, for example, to expand their descriptions or provide some initial feedback on the perceived usefulness of the concepts generated.

2.3. Data Management and Analysis

Robust data security measures were defined to ensure that the compilation of personal data from the recruitment process and during the co-creation event complied with the General Data Protection Regulation (GDPR) as well as to ensure the integrity of the data. These practices consisted of an informed consent form collected from all participants ahead of the co-creation activity, the use of a password to access raw data compiled in the recruitment questionnaire, the application of pseudonymization techniques to replace contact details, and the storage of pseudonymized data in the Move2CCAM consortium SharePoint and in a data warehouse from the project.

Once the data collection process concluded, the dataset acquired from the recruitment questionnaire and the online platform was unified in an Excel file through the ID number. An iterative analysis process was used to identify a set of variables that defined the different mobility services suggested for autonomous vehicles (e.g., the type of vehicle, trip purpose, frequency, target user, etc.). Data were coded independently by two researchers to reduce the risk of bias, with researchers then comparing notes to ensure a common understanding of the ideas provided by the participants. The researchers ensured each participant had a full set of responses to ensure that all contextual information provided was considered (e.g., where an idea was suggested in phase 2 and developed in phase 3); however, a small number of ideas were insufficiently clear, and were not included in the analysis.

To enable a comparison of co-created ideas between demographic groups, we further analyzed the coded data to group similar concepts into archetypes, by which we mean broad categories of vehicle and service types. A comparison of the ideas generated across demographic groups was performed at the European level and not at the country level to maintain sufficient sample sizes for comparison. More details on the Materials and Methods used can be found in [27].

3. Results

3.1. Analysis of Co-Created Ideas

Through the data analysis process described above, 337 scenarios for autonomous vehicles were identified. These scenarios could imply widely used services (e.g., a public vehicle covering a generic route) or specific services (e.g., a vehicle adapted to a particular type of traveler and serving a specific location or for a specific purpose).

The key differential variables considered for the definition of scenarios were the following:

• Vehicle typology;
• Vehicle size;
• Energy source;
• Target users;
• Trip purpose;
• Locations served;
• Distances traveled;
• Frequency;
• Ownership;
• Safety measures;
• Accessibility features;
• Methods of payment.

By grouping the most similar ideas together, seven archetypes were identified, which encompassed 80% of the ideas generated, with the remaining 20% describing a diversity of use cases, which were suggested by fewer than eight participants each.

The archetypes identified are described below.

1.

Passenger transport use cases

• An autonomous e-hailing shared pod refers to an autonomous vehicle with a capacity for between two and eight passengers that covers short and medium distances (i.e., up to 30 km) and provides a door-to-door service shared between different users. Participants typically described this as an on-demand mobility service hired by users through an app operated by a private organization. This case is similar to current car-sharing services but without a driver.
• An autonomous private car refers to a small vehicle whose owner can use it according to their mobility needs. This is just like a current private car but without the need for a driver.
• An autonomous bus consists of a vehicle with a capacity for 10–60 passengers that provides a service with a fixed or flexible timetable and routes. Suggestions provided by participants included a variety of distances traveled, destinations served, and trip purposes and encompassed public and private operators.

2.

Freight transport use cases

• Delivery robots are small-sized autonomous surface vehicles that transport freight over short distances, for example, packages delivered to homes. Participants often suggested vehicles operating outside of peak hours and specific examples like delivering medicines or deliveries within large facilities like a factory.
• Delivery by vans includes medium-sized autonomous vehicles to transport small, medium, and sporadically large freight to medium distances (i.e., up to 15 km) working on demand and operated by a private organization.
• Delivery drones consist of small aerial autonomous vehicles delivering small freight over short distances and operating on-demand. The type of freight to transport includes food, medicines, and small packages, and the area of coverage could be large or part of a facility, such as factories.
• Autonomous long-distance trucks refer to large-sized autonomous vehicles that transport medium and large goods over long distances through optimized supply chains to ensure timely and reliable freight transportation. Examples provided by participants include all kinds of trucks delivering freight by road, including platooning trucks.

3.

Other.

• This group involves types of vehicles or services suggested by only a small number of participants (less than eight). Many of the ideas collected here involve autonomous vehicles for specific functions such as emergencies, farming, construction, military use, garbage trucks, and platooning pods. This category includes both passenger and freight vehicles.

3.2. The Future That Citizens Imagine Is Dominated by Shared, Electric, Passenger Vehicles

Exploring the number of times a particular archetype was suggested by participants in the co-creation process, as per Figure 1, reveals how widespread expectations are for these use cases among participants.

• Passenger transport was the most co-created solution. Of the 337 ideas generated, 218 were for passenger services (over two-thirds) compared with 52 for freight transport. There was a high degree of consistency in the use cases proposed for passenger transport, with autonomous buses and e-hailing shared pods being by far the most common.
• Freight solutions were much less frequently suggested. When participants mentioned freight, suggestions were much more variable, covering drones, robots, vans, and long-distance vehicles, among others.
• Shared mobility was the norm as, of the 218 passenger services proposed, the large majority were shared (buses or pods) rather than private vehicles, which were suggested just 40 times.
• AVs were assumed to be electric, with the exception of long-distance freight transport, which was considered powered by hydrogen, and participants consistently reported that vehicles would be electric.
• The archetypes were largely differentiated by vehicle type and size, the distance traveled, and ownership. Other variables like target users, methods of payment, and frequency were highly variable within each use case.

Appendix A includes more information about the number of AV co-created by each type of participant.

3.3. Indicative Analysis of Co-Created Ideas by Demographic Group

To analyze the relationship between demographic characteristics and perceptions of AVs, the demographic profiles of those suggesting passenger and freight vehicles were compared with the sample as a whole. Six demographic characteristics were included in this analysis, which correspond with the following:

• Socio-demographic aspects that cover age, gender, and income.
• Mobility needs and behavior associated with health conditions, driving capacity, and place of residence.

Given the small sample size of the co-creation study, these findings are indicative; however, they do offer novel insights into how different demographic groups may imagine the future. Charts have been provided to visualize the reported differences; however, these do not represent statistical analysis.

3.3.1. Who Is Imagining Autonomous Passenger Vehicles?

Looking across the three main categories of passenger vehicles (e-hailing shared pods, private cars, and buses), ideas were suggested by participants with a range of demographic criteria, largely consistent with the sample as a whole, as shown in Figure 2. These use cases were suggested frequently by participants in our study, indicating they are widely known and are seen to have broad applicability across age and gender boundaries in a range of urban and rural settings in all countries included in the sample and for those with a range of mobility needs.

Looking at specific use cases, minor differences occur in the patterns of users suggesting different approaches. For example, e-hailing services are slightly more frequently suggested by people in middle- or low-income categories; however, across other categories, including urban/rural locations, there are no strong trends. As with the broader category of passenger vehicles, the e-hailing shared pod is part of the future envisaged by a wide range of people in different demographic categories and mobility situations.

For private autonomous cars, there were slightly more suggestions from younger people (who were less likely to hold driver’s licenses), those in higher income brackets (who are more likely to own a private car currently), and those in suburban and town settings (where public transport may be less available). These indicative findings suggest that the personal relevance of transport mode does have some influence on the futures people imagine, but not overwhelmingly so. Similar patterns were identified in the suggestions for autonomous buses, which are suggested to be used slightly more by women, those with health problems, those in small towns and suburbs, and all groups who may particularly benefit from improved bus services.

3.3.2. Who Is Imagining Autonomous Freight Vehicles?

As noted above, fewer citizens made suggestions related to freight transport uses, suggesting that they are a less well-established part of the shared imaginary of self-driving vehicles. Across the four main categories of freight use cases (delivery robots, drones, vans, and long-distance trucks), these were more frequently suggested by male participants, in contrast to the even split of men and women, suggesting passenger use cases (as shown in Figure 3). This finding may simply reflect the greater likelihood of male participants working in the male-dominated logistics industry, giving freight uses greater salience, rather than women who might not value freight services becoming autonomous.

4. Discussion

The findings presented here represent a novel source of insight into public expectations, hopes, and aspirations for autonomous vehicles and services. It brings together data from five EU countries, a more diverse and heterogenous sample of citizens than is typical, and presents their spontaneous visions for the future of self-driving vehicles rather than their reactions to expert-derived scenarios. Here, the significance of these findings, particularly in relation to other data sources on public attitudes, is explored.

4.1. There Is a Consistent Shared Vision for Autonomous Passenger Transport Across Diverse European Countries

Analysis shows that in all five locations, and for a wide and diverse selection of citizens, automation is seen to offer useful transport solutions for passengers. Citizens generated over 300 different ideas for how automated vehicles and services could be deployed, of which two-thirds fell into three broad archetypes of passenger transport. Autonomous buses and e-hailing pods, in particular, are novel modes that appear to be part of a broadly shared understanding of the future of autonomous vehicles. This understanding seems to span beyond those who see themselves using these services immediately. In contrast with stated preference surveys, which find variation in their willingness to use autonomous vehicles among different demographic groups, our study found that both shared and autonomous services are envisaged by a wide range of individuals.

In contrast, there was little consensus among participants about the future of autonomous freight services, with many fewer and more varied suggestions made. The implications of this absence of consensus may be positive—leaving space for positive disruption—or negative, suggesting a number of individuals who are less accepting of new technological solutions that feel unfamiliar.

4.2. Citizens’ Imagined Transport Systems with Automation Encompassing Electric and Shared Mobility

Looking across all demographic groups, co-creation data showed a shared vision of a transport system that is very different from today. Firstly, citizens imagined that all autonomous transport would use sustainable fuel—either electric or hydrogen—showing how widely accepted the move to sustainable fuels is across Europe. Secondly, there are many more suggestions for shared mobility services than for private vehicles, which is a reversal of current transport patterns where private cars account for the majority of trips across Europe. These findings suggest a greater willingness from citizens to adopt shared services than studies assessing the instrumental value of travel time or the affective and symbolic value of vehicle ownership (as per a recent literature review).

A useful future development would be to explore more explicitly how the expectation of participants, as expressed in the co-creation setting, may differ from their likely behavior—providing further explanation of the differences between studies such as ours and others that challenge the assumption of greater electric and shared mobility on the basis of stated preference studies.

4.3. How Citizens Imagine This Transport System Operating

Moving forward from the specification of the vehicle or service to the descriptions citizens used, the analysis of AV services provided by participants indicates how they see this new transport system operating. While participants in the co-creation activity did mention the value of reallocating time spent driving, they more frequently drew out other benefits over current transport systems. Participants commonly described new services as more frequent, more convenient, or more flexible than existing transport options, either because of extended schedules, door-to-door pickups, or demands for responsive scheduling. Many of the ideas co-created included reference to specific groups who would benefit from greater access to mobility, particularly those with disabilities, health conditions, or age-related issues that prevent them from driving. Participants often described autonomous services as more cost-effective than private transport, although perhaps more expensive than current public transport options, which is the price of greater convenience and flexibility.

These findings suggest that automated vehicles are largely valued for the benefits they can bring above and beyond the lack of a driver. This should inform the design of services and policy to support the autonomous vehicle rollout, which should seek to maximize wider benefits. For vehicle designers and manufacturers, these findings confirm the importance of accessibility if the desired goal of greater inclusion is to be met.

4.4. Who Is Imagining Which Future: The Risks of Replicating Inequalities

While our analysis of demographic differences is limited by the small sample size of the study, the high level of diversity and careful stratification relative to many larger convenience sample studies does allow us to draw out indicative findings. Looking across the use cases suggested by participants, it is notable that passenger transport use cases (e-hailing, private vehicles, and buses) are suggested by men and women, by people in a range of income brackets, with and without health conditions, and in a range of urban and rural settings across the five European countries of our study. This strongly suggests that these use cases are widely known, accepted, and seen as applicable in diverse settings.

However, there are differences that seem to correlate with current trends. For example, autonomous private cars were suggested more frequently by people in the high-income category and those without health problems, characteristics of which are both correlated with each other (e.g., people living with ill health and disabilities are disproportionately likely to live in poverty) and negatively correlated with car ownership. This effect is replicated with the autonomous bus service, which was suggested more often by women and people with health problems who are less likely to drive. This illustrates the potential for AV services to replicate existing patterns of transport usage, which in many cases exacerbates inequality—without active intervention to increase accessibility, there is a strong likelihood of replicating existing problems.

There are also some observable differences according to the place of residence—with both autonomous private cars, and autonomous buses suggested more often by those living in suburban and town settings, where public transport is typically less prevalent. This speaks to a perceived opportunity for autonomous services to fill existing transport gaps, adding value over and above the potential benefits of switching from existing modes.

4.5. Implications for Future Research

This study covers several regions of different European countries and includes the collaboration of a small but inclusive sample of citizenships. Working on this scale and using the open-ended and creative co-creation method allowed us to obtain a more detailed description of the mobility services imagined by citizens. It stimulated thinking both among research participants and researchers by bringing together diverse perspectives, leading to richer research findings. We encourage others to continue using co-creation activities to explore societal transformations like the move to autonomous vehicles for these reasons. They add a valuable and distinct perspective compared to more instrumentally focused studies such as stated preference surveys and other forms of opinion research.

However, future studies with a higher number of participants are required to enable a more robust characterization of the profile of citizens reporting each of the scenarios co-created. Future research on these directions could also explore more directly the difference between the desires and expectations of citizens in relation to autonomous vehicles.

5. Conclusions

With this pan-European co-creation activity, we set out to understand the kinds of futures citizens envisage for autonomous vehicles. In contrast with the existing literature on preferences and acceptability, it did not prescribe to participants what services were possible but invited them to describe their own visions of the future. We provide a detailed account of this methodology and encourage its wider use to further broaden our understanding of citizens’ imagined futures.

This study found a high level of consistency across all five EU countries when it comes to the role of autonomous vehicles in passenger transport. Participants readily identified autonomous buses, e-hailing pods, and, to a lesser extent, private autonomous cars as key components of a future transport system despite their individual demographic characteristics. This suggests a base level of acceptance of these modes, even among those who are unlikely to be early adopters.

Going beyond vehicle type, we also found that autonomous vehicles were consistently imagined to be both electric and shared. While electrification is broadly seen as a fixed component of a future transport system in Europe, debates continue in research and policy about whether a shift to shared mobility is plausible. This finding suggests that the concept of shared mobility has entered the public consciousness and is strongly associated with autonomous vehicles, providing a basis from which modal shifts may emerge.

Finally, while we found a high degree of consistency in some elements of the way the public envisages autonomous vehicles, other aspects are less well defined—such as freight transport—which is likely to hinder public acceptance. We also identified indicative trends for individuals who envision futures that replicate current transport patterns, which in many cases are both unsustainable and inequitable. We encourage transport planners and autonomous vehicle developers to consider carefully how choices at this formative stage may reproduce or challenge these inequalities.