1. Introduction
As urban mobility systems transition toward automation, public acceptance emerges as a central challenge for policymakers and operators. Technological readiness alone does not ensure societal integration; rather, long-term success depends on aligning technological capabilities with user expectations.
In Germany, pilot projects for autonomous buses – e.g., HEAT (Hamburg Electric Autonomous Transportation), KIRA (AI-based regular operation of autonomous on-demand transport), MINGA (Munich’s automated local transport with ride pooling, solo buses, and bus platoons) – are expanding rapidly, yet public acceptance remains heterogeneous (VDV, 2025). While some groups welcome automation and sustainable transport innovations, others express skepticism regarding safety, comfort, and reliability (Reference AlqahtaniAlqahtani, 2025).
Segmenting the population into socio-psychologically coherent user groups enables anticipation of behavioral responses, identification of adoption barriers, and development of tailored acceptance strategies. This study investigates the latent structure of mobility users in the Munich Metropolitan Region to determine who is most likely to adopt autonomous public transport and under what conditions.
Existing research on autonomous public transport often focuses on aggregate attitudes or single determinants such as trust, perceived risk, or technology readiness. Such approaches overlook the underlying heterogeneity of value systems and lifestyle-based mobility preferences. The key gap lies in the absence of empirically grounded user typologies that link behavioral preferences (e.g., speed, cost, comfort, environmental consciousness) to openness regarding autonomous bus usage. (Reference Chatziathanasiou, Botzoris, Morfoulaki and KotoulaChatziathanasiou et al., 2024; Reference Golbabaei, Yigitcanlar, Paz and BunkerGolbabaei et al., 2022; Reference Öztürker, Homem de Almeida Correia, Scheltes, Olde Kalter and van AremÖztürker et al., 2022)
Without such typologies, policy recommendations risk remaining generic and ineffective. A systematic segmentation based on behavioral and attitudinal dimensions is therefore essential for explaining not only how many people would use autonomous buses, but which types of users would do so and why.
The objective of this research is to identify and characterize clusters of mobility users in the Munich Metropolitan Region using preference-based and socio-demographic indicators from a large-scale online survey. Through a multidimensional clustering approach, the study aims to:
-
1. Identify latent groups that differ in their subjective valuation of core mobility dimensions (speed, cost, environmental friendliness, comfort, safety, reliability).
-
2. Examine how these clusters vary in their perceived utility (subjective expected utility, SEU) of conventional and autonomous transport modes.
-
3. Discuss the implications of these findings for the design and communication of autonomous public transport systems.
2. State of the art
The analysis of heterogeneous mobility behavior through cluster segmentation has become an established methodological and conceptual approach in transport research. In contrast to classical transport modeling, which often treats the population as a homogeneous group of rational actors, cluster-based analyses acknowledge that mobility behavior is embedded within broader social, psychological, and value-based contexts (Reference StrömbladStrömblad, 2024). From a behavioral transportation perspective, individuals’ mobility choices are influenced not only by objective mode attributes such as cost or travel time but also by subjective utility structures, personal lifestyles, and prevailing social norms (Reference Bhagat-Conway, Mirtich, Salon, Harness, Consalvo and HongBhagat-Conway et al., 2022). Consequently, attitudes toward mobility reflect instrumental, affective, and symbolic motivations. Segmentation techniques allow researchers to uncover latent user groups that differ in their motivational priorities, offering a more nuanced understanding of potential acceptance patterns for innovations such as autonomous buses.
While early segmentation approaches primarily relied on socio-demographic variables such as age, gender, or income, contemporary mobility research increasingly employs psychographic clustering based on attitudinal, behavioral, and normative indicators to identify distinct mobility mindsets. (Reference Chatziathanasiou, Botzoris, Morfoulaki and KotoulaChatziathanasiou et al., 2024; Reference Magdolen, Behren, Burger and ChlondMagdolen et al., 2021; Reference Vallée, Ecke, Barthelmes and VortischVallée et al., 2024) Such clusters capture how individuals conceptualize mobility in relation to their identities and daily routines, providing deeper insights into mode-shift willingness and the adoption of new technologies.
Beyond descriptive segmentation, integrative theoretical models such as the Unified Theory of Acceptance and Use of Technology (UTAUT2) (Reference Venkatesh, Brown and BalaVenkatesh et al., 2013) and its derivatives (Reference Golbabaei, Yigitcanlar, Paz and BunkerGolbabaei et al., 2022; Reference Tamilmani, Rana and DwivediTamilmani et al., 2021) establish links between cluster membership and behavioral intention through constructs including performance expectancy, effort expectancy, social influence, and hedonic motivation. However, these frameworks typically assume linear relationships and homogeneous populations. Cluster-based analyses complement such models by empirically revealing non-linear and group-specific relationships, thereby enhancing the explanatory power of acceptance research.
Recent studies on autonomous public transport (Reference Chahine, Christ and GkritzaChahine et al., 2024; Reference Hong, Park and LimHong et al., 2025; Reference Korkmaz, Fidanoglu, Ozcelik and OkumusKorkmaz et al., 2021; Reference Molinillo, Caballero-Galeote, Liébana-Cabanillas and Ruiz-MontañezMolinillo et al., 2024; Reference Wu, Zhou, Xi and WuWu et al., 2021) have shown that public perceptions of automation are strongly shaped by trust, perceived safety, and compatibility with personal values. Yet these investigations often fall short of identifying which population segments are most likely to adopt such systems. This gap highlights the absence of empirically grounded user typologies that connect individual preferences with adoption potential.
Positioned within this evolving field, the present study extends the cluster-based mobility research paradigm to the context of autonomous bus systems. Building on established segmentation approaches and decision theory foundations, specifically the concept of subjective expected utility (SEU) (Reference EsserEsser, 1993), it integrates value orientations and probabilistic perceptions of mode performance to better predict adoption patterns under real-world conditions. (Reference Hassanpour, Costa and Oliveira CruzHassanpour et al., 2025; Reference Welte-BardtholdtWelte-Bardtholdt, 2022) For example, a SEU approach has already been used to reliably predict the choice of more traditional transport modes, such as cars, bikes and public transport (Reference Weyer and HoffmannWeyer & Hoffmann, 2025). By leveraging these theoretical and empirical foundations, this research contributes to the next generation of mobility segmentation by explicitly linking behavioral typologies to technology-specific acceptance scenarios. In doing so, it enables policymakers and planners to design differentiated engagement and service strategies that align with the psychological diversity of future autonomous bus users. (Reference Ahmed, Hizam and SentosaAhmed et al., 2022; Reference Chatziathanasiou, Botzoris, Morfoulaki and KotoulaChatziathanasiou et al., 2024)
3. Methodology
In the following section, we present the methodological approach used to identify the distinct mobility clusters as determined in this study.
3.1. Research design and conceptual approach
The present study adopts a quantitative, exploratory, and cross-sectional research design aimed at identifying and characterizing latent mobility user segments in the Munich Metropolitan Region. The underlying assumption guiding the design is that individuals act according to the principle of SEU (Reference EsserEsser, 1993). They select transport modes that maximize the anticipated utility derived from their personal goals, weighted by their perceived likelihood that these goals will be achieved through specific mobility options. Accordingly, the methodological framework integrates two analytical components:
-
1. Measurement of subjective mobility preferences and expected mode performance, and
-
2. Identification user groups based on these multi-dimensional preference profiles via cluster analysis.
This dual approach bridges the gap between psychological attitude research and quantitative mobility modeling, enabling a deeper understanding of who is likely to use autonomous public transport under which motivational and contextual conditions. (Reference Weyer and HoffmannWeyer & Hoffmann, 2025)
3.2. Data collection and sample characteristics
The data were obtained from a large-scale online survey conducted between August and December 2024. Participants were recruited through a combination of targeted online advertisements, public transport mailing lists, community newsletters, and a panel service provider.
A total of 1,755 respondents completed the survey. After applying rigorous data quality filters, such as the exclusion of incomplete responses and implausible completion times, the final sample consisted of 1,540 valid cases. Approximately 77% of participants reported residence within Bavaria, while 23% originated from other German regions. Preliminary analyses found no significant differences between these groups in terms of mobility preferences or attitudes toward automation, justifying their inclusion in the main dataset to preserve statistical power.
Socio-demographic breakdown:
-
• Gender: 57% female, 42% male, and 1% non-binary.
-
• Age range: 18-75 years (M = 42.6, SD = 12.7).
-
• Education: 62% high school, 29% secondary school, 6% primary school, 1% no education, and 2% did not respond.
-
• Transport resources: 78% access to private car, 71% own bicycle, 65% hold public transport subscription.
3.3. Measurement constructs
The questionnaire measured respondents’ subjective mobility preferences, mode-specific performance perceptions, and attitudes toward autonomous public transport.
3.3.1. Mobility preference dimensions
As previous studies indicate (Reference SchneiderSchneider, 2013; Reference Wolf, Schröder, Neumann and HaanWolf et al., 2015), transport mode choices for routine travel purposes are shaped by a range of personal, mobility-related preferences (i.e., goals, needs, value orientations, and similar concepts). Respondents rated the importance of six core mobility attributes on a 0-10 scale, representing their preferences (U-values) when choosing transport modes:
-
1. Speed – perceived time efficiency and directness of travel,
-
2. Cost efficiency – perceived affordability and value for money,
-
3. Environmental friendliness – alignment with sustainability values,
-
4. Comfort – convenience, seating quality, and travel atmosphere,
-
5. Safety – both objective and perceived protection from accidents or incidents, and
-
6. Reliability – punctuality and predictability of travel times.
An additional variable, “stress-free travel,” was collected to assess potential characteristics of autonomous systems but was not included in the cluster formation so as to maintain comparability with traditional modes and comparable studies.
3.3.2. Subjective expected utility (SEU) estimation
For each transport mode (private car, public transport via bus or train, bicycle, and autonomous bus), respondents evaluated how well the mode fulfilled each of the six preference dimensions from their perspective. These perceived individual probabilities (p-values) were multiplied by the respective importance weight of each preference (U-value) using the equation (1) presented below to form an overall SEU (Reference EsserEsser, 1993) score per mode for each participant.
SEU denotes an individual’s subjective expected utility of a transport mode i, taking into account the six core mobility attributes j. This operationalization captures how individuals internally evaluate trade-offs between their mobility values (e.g., “comfort versus sustainability”) and mode characteristics (e.g., “autonomous buses are reliable but not yet comfortable”).
3.4. Cluster analysis procedure
-
1. Preprocessing, outlier detecting and initial clustering:
Prior to clustering, the responses were examined for the previously described patterns and instances of illogical response behavior, which were subsequently excluded from the analysis. In all typification attempts, outliers were identified using the single-linkage method, after which eight cluster solutions with varying numbers of clusters (i.e., all 2- to 9-cluster solutions) were generated using Ward’s method.
-
2. Partitioning and validation:
In order to find a stable cluster solution, the mean values of the initial Ward clusters were then used as input for a subsequent clustering based on the K-means method, which represents a non-hierarchical optimization algorithm that reassigns observations iteratively to minimize within-group variance. In this context, a cluster solution is stable if the majority of people is reliably assigned to the same cluster in both methods. In addition to other statistical and visual criteria (e.g., variance ratio criterion, silhouette coefficient, dendrogram plot), primary emphasis was placed on ensuring that the selected cluster solution allowed for clusters that were conceptually coherent and logically interpretable. Reference Sarstedt and Mooi(Sarstedt & Mooi, 2019)
After comparing interpretability, validity, and cluster size balance, a five-cluster solution was selected as the most comprehensible.
4. Results
The final analysis yielded a five-cluster solution that offered the most interpretable and statistically stable representation of heterogeneous mobility preferences among respondents. The five clusters captured distinct behavioral and attitudinal profiles reflecting trade-offs between functional, emotional, and normative mobility values. Cluster sizes range from 12% to 30% of the sample.
Cluster 1 – Risk-averse environmentalists (26%)
Cluster 1 encompasses individuals who place a strong normative emphasis on environmental protection. This group demonstrates eco-centric rationality. They consciously trade speed and comfort for sustainability (+47.2%) and to a much lesser extend for safety (+5.4%). Modal split data indicate frequent public transport and cycling (23%) usage, with minimal private-car dependency. Socio-demographically, this cluster tends to be older – 24.2% are above 60 years old, compared to 19.7% of the total sample; and only 25.2% are under 30 years old, in comparison to 29.7% total. Regarding autonomous mobility, cluster 1 respondents exhibit high openness.
Figure 1 illustrates the deviations of the six individual preferences from the sample mean across the five identified clusters.
Cluster 2 – Pragmatists (17%)
Mobility preference for each cluster with average scores and deviation
Figure 1 Long description
The bar graph compares mobility preferences across five different user clusters: Risk-averse environmentalists, Pragmatists, Indifferent moderates, Comfort-oriented car users, and Cost- and environmentally conscious users. The horizontal axis lists these clusters, while the vertical axis represents the percentage deviation from a baseline, ranging from -80 percent to 40 percent. Each cluster has six vertical bars representing different mobility preferences: Speed, Cost efficiency, Environmental friendliness, Comfort, Safety, and Reliability. The colors of the bars are as follows: Speed is dark blue, Cost efficiency is medium blue, Environmental friendliness is light blue, Comfort is dark teal, Safety is medium teal, and Reliability is light teal. Key observations include: Cluster 1 shows a high preference for Environmental friendliness at 47.2 percent and significant dispreference for Speed at -19.1 percent. Cluster 2 has a notable dispreference for Cost efficiency at -75.5 percent. Cluster 3 exhibits a balanced preference with slight variations. Cluster 4 shows a high preference for Safety at 22.1 percent and significant dispreference for Speed and Cost efficiency, both around -28 percent. Cluster 5 has a high preference for Cost efficiency at 21.9 percent and a significant dispreference for Reliability at -35.2 percent.
Cluster 2 users highly emphasize nearly all the preferences, such as efficiency and a fast, inexpensive and safe transport. Environmental friendliness is the only aspect that is not a concern for this user group. This utilitarian orientation translates into a car-centered lifestyle. Roughly 34% report the private car as their dominant mode. Respondents frequently expressed doubts about technical reliability and loss of control, framing automation as a potential constraint on autonomy rather than an improvement. Socio-demographically, cluster 2 includes more male full-time employees in the 30-50 age range, often with long commutes. Their mobility pattern reflects a performance-driven logic: time efficiency and predictability outweigh environmental or social considerations.
Cluster 3 – Indifferent moderates (30%)
This is the largest cluster and displays a moderate overall preference profile featuring smaller standard-deviation shifts from the sample mean across all six attributes. Only speed and comfort exhibit notable deviations in this group, with differences of approximately -10% and 20% respectively.
The lack of strong preference polarization suggests a adaptive mobility style. Members use whichever mode is situationally convenient. Sociologically, cluster 3 can be regarded as a “latent adopter group.” They are neither early enthusiasts nor rejecters but could shift behaviorally if autonomous services become visible, socially normalized, and integrated into everyday routines.
Cluster 4 – Comfort-oriented car users (12%)
Cluster 4 is small, comprising only 12% of the users, but nonetheless distinctive. Members prioritize speed (22.1%) and comfort (16.7%), while showing a disinterest in environmental (-28.3%) and cost considerations (-28.1%). Over 37% of respondents in this cluster identify the private car as their main transport mode. Their mental availability of alternative transportation modes is low (-16%). This suggests that for cluster 4, perceived service quality is a critical adoption lever.
Cluster 5 – Cost- and environmentally conscious users (15%)
Cluster 5 resembles cluster 2 in some aspects, as speed (14.7%) and cost efficiency (21.9%) are also rated above average. However, comfort and safety are weighted substantially lower at -35.2% and -23.0%. Most importantly, the preference for an environmentally friendly transportation mode – at 17.0% above average – deviates heavily from cluster 2’s -75.5%. These pronounced differences across three key dimensions justify distinguishing between clusters 2 and 5, despite their similarities in the attributes mentioned.
Cluster 5 features a lot of people who use a bicycle as their transportation mode and education levels are high. This reflects their alignment between pro-environmental and cost-sensitive motives. Respondents perceive automation as an efficient means of reducing congestion and emissions, provided that fares remain low and integration with existing networks is seamless.
4.1. Comparative analysis of modal utility
A variance analysis of SEU scores across the four transport modes (private car, public transport, bicycle, autonomous bus) yielded significant inter-cluster differences, as shown in Figure 2 below:
-
• Clusters 1 and 5 assign higher SEUs to collective and sustainable modes (autonomous bus, cycling) and lower SEUs to the car,
-
• Clusters 2 and 4 maintain strong attachment to individual motorized transport, and
-
• Cluster 3 occupies an intermediate position, with moderate scores across all modes.
Deviations of the average cluster SEU values from the overall mean (values in brackets)
4.2. Implications for autonomous public transport
The clusters reveal how individuals exhibit different priorities that must be satisfied by an autonomous bus in order to ensure successful implementation.
Cluster 1 – Risk-averse environmentalists:
Likely early adopters. Strategies should emphasize ecological benefits (e.g., zero-emission operation or energy-efficiency), and safety assurance (e.g., transparency about safety standards, data security or emergency systems). Demonstration projects could use this group as ambassadors for sustainable innovation. Due to safety concerns and the slightly higher age of cluster members, such projects should be supervised by support staff and contact persons, and vehicles should be designed to be barrier-free.
Cluster 2 – Pragmatists:
Require time for behavioural change and strong functional incentives: speed, punctuality, and cost efficiency must clearly outperform the private car. Policies such as dedicated autonomous vehicle lanes or fare subsidies may increase perceived utility. Additionally, due to a strong dependence on private cars (e.g., for commuting) and the resulting potential inertia in mobility behavior, autonomous buses should be integrated into existing commuter routes or park-and-ride systems to provide a low-threshold introduction and positive initial experiences.
Cluster 3 – Indifferent moderates:
Represent a latent mass market. Their adoption hinges on social diffusion and habitualization (e.g., intuitive services or visibility of autonomous vehicles in media and everyday life). Communication should focus on reliability, normality, and convenience rather than technology.
Cluster 4 – Comfort-oriented car users:
Adoption possible only if autonomous buses offer premium service quality. Due to reduced mental availability and high dependence on cars, the entire driving experience should be designed to feel like a “normal” car trip rather than a trip with public transport: Integration with on-demand shuttles (instead of stop-based services) or “executive” comfort features (e.g., comfortable, adjustable seats or personal control over temperature) could attract this segment.
Cluster 5 – Cost- and environmentally conscious users:
Represent value-driven early adopters. Policies and communication should underline affordability (e.g., by including trips with autonomous buses in existing public transport tickets or services), energy efficiency, and environmental impact reduction (e.g., displaying ecological footprint per trip or providing options to combine autonomous buses with bikes for multimodal trips). Autonomous buses could thus be integrated into existing public transport services.
4.3. Synthesis
In synthesis, the cluster analysis demonstrates that mobility behavior and acceptance of autonomous public transport are multi-causal phenomena influenced by both normative and instrumental value orientations. Rather than a single continuum of acceptance, the results reveal a segmented adoption landscape: Two clusters (1 and 5) show high readiness, aligning with sustainability-oriented innovation narratives. One cluster (3) remains neutral but convertible through targeted exposure and information. Two clusters (2 and 4) exhibit structural resistance, demanding strategic intervention in terms of service performance and image.
The diversity of motivational drivers implies that successful deployment of autonomous buses will depend on adaptive governance, flexible service models, and differentiated communication attuned to distinct user psychologies.
5. Discussion
5.1. Interpretation of the cluster structure
The five-cluster typology identified in this study provides compelling evidence that mobility preferences in the Munich Metropolitan Region are heterogeneous and value-driven. Each cluster represents a distinct motivational constellation that translates into differential acceptance patterns for autonomous public transport.
The existence of both eco-rational (clusters 1 and 5) and performance-rational (clusters 2 and 4) user types highlights the dual logic that underlies mobility behavior in contemporary societies. On one side, mobility is perceived as a moral and ecological act that represents a contribution to sustainability and social responsibility. On the other, it remains an expression of autonomy, control, and efficiency.
Cluster 1 and 5 correspond to early adopter categories, comprising individuals with positive attitudes, strong normative alignment with sustainability, and openness to innovation. In contrast, clusters 2 and 4 embody late majority or laggard profiles, marked by instrumental motivations and higher perceived behavioral control barriers.
The indifferent moderates (cluster 3) occupy a pivotal position in this diffusion process. Their neutrality implies neither strong resistance nor proactive interest. However, as diffusion theory suggests, the mass adoption phase of any new technology depends on such undecided majorities becoming habitual users once systems gain social visibility and normative legitimacy Reference Rogers(Rogers, 2003).
The cluster typology therefore offers a dynamic framework for understanding how different user groups – from early ecological adopters to pragmatic mainstream users – may successively enter the adoption curve of autonomous public transport.
5.2. The role of values and attitudes
The present findings reinforce the centrality of values as mediating variables between socio-demographic characteristics and behavioral intentions. Consistent with prior work Reference Golbabaei, Yigitcanlar, Paz and Bunker(Golbabaei et al., 2022), value orientations toward environmental sustainability, safety, and comfort emerge as stable psychological anchors shaping transport-related decision-making.
Clusters 1 and 5 demonstrate that value-behavior congruence – alignment between environmental values and mobility choices – strongly predicts openness to autonomous buses. These respondents perceive automation as an enabler of collective sustainability rather than an abstract technological novelty.
Conversely, clusters 2 and 4 reveal value-behavior dissonance. Although these individuals may express awareness of sustainability issues, their daily routines prioritize autonomy, speed, and comfort. Such dissonance explains why information campaigns emphasizing environmental benefits alone often fail to alter mobility behavior Reference Anable(Anable, 2005). For these groups, instrumental attributes (e.g., punctuality, convenience, and control) must be improved before normative appeals become persuasive.
Importantly, cluster 4 underscores the emotional dimension of mobility. Comfort and privacy are not mere functional attributes but expressions of psychological security. This insight aligns with findings from qualitative studies on car attachment Reference Kent(Kent, 2015), suggesting that resistance to automation often reflects identity-related concerns rather than rational skepticism.
5.3. Implications for acceptance of autonomous public transport
The cluster typology has several implications for the strategic design, implementation, and communication of autonomous public transport systems, such as autonomous buses:
Differentiated communication strategies:
Public messaging must be differentiated. For eco-rational clusters, communication should emphasize environmental benefits, system efficiency, and contribution to the energy transition. For performance-rational clusters, messages should stress technological reliability, time savings, and convenience – framing automation as an enhancement, not a restriction.
Service design customization:
Autonomous bus systems should reflect the heterogeneity of user expectations as shown in Figure 1. As an example for cluster 5, affordability and ecological transparency could strengthen identification.
Policy integration and incentives:
Local governments can leverage cluster insights to design targeted incentives. For performance-oriented users, travel-time advantages (e.g., dedicated autonomous vehicle lanes) or priority signalization could compensate for initial skepticism. For environmentally conscious groups, fare integration with sustainable mobility modes could foster multimodal acceptance.
Pilot project composition:
The findings caution against recruiting only “tech-savvy enthusiasts” for pilot programs. To achieve representative feedback, field trials should intentionally include participants from resistant clusters to uncover latent usability and trust barriers early in the design process.
Social norm formation:
Since cluster 3 represents a neutral mainstream, large-scale adoption may depend on social diffusion mechanisms. Seeing peers or respected community members use autonomous buses can normalize the technology. Hence, public demonstrations and positive media coverage may play a catalytic role in shifting social norms.
Socio-psychological mechanisms of acceptance
The results suggest that acceptance of autonomous public transport operates through three interlinked psychological mechanisms:
-
1. Instrumental evaluation: Users compare perceived functional performance (speed, safety, reliability) against existing transport options. This mechanism dominates clusters 2 and 4.
-
2. Affective association: Emotional comfort, perceived control, and trust in technology shape intuitive responses, with particular relevance for clusters 3 and 4.
-
3. Normative identification: Alignment with moral or collective goals (e.g., sustainability, equity) underpins acceptance among clusters 1 and 5.
The study illustrates that the relative weight of these components differs systematically across clusters, implying that interventions must be psychologically tailored rather than uniformly applied.
5.4. Methodological reflection
The study’s use of SEU combined with cluster analysis provides a novel methodological lens. Nonetheless, the applied methods and our findings are subject to certain limitations. First, the cross-sectional design captures intentions and stated preferences, which may deviate from actual behaviour. Similarly, the SEU of autonomous buses was surveyed in very broad terms. Given the lack of user experiences and behavioural data in this area, gathering additional information on potential usage contexts or scenarios could be helpful in providing deeper insights into acceptance patterns (e.g., commuting, leisure, time of travel, presence of other passengers). Second, the online sample may overrepresent digitally literate populations. Last, regarding the cluster analysis, moderate silhouette coefficients indicated a partial overlap between user profiles, which is common in attitudinal data but limits cluster exclusivity. One way to illustrate the clusters more clearly would be to design ‘personas’ based on the present findings, i.e., fictional single individuals who are representative of a cluster. Future research should thus employ mixed-method triangulation, combining quantitative segmentation with qualitative interviews, to validate cluster interpretations and explore underlying motivations, needs, and problems more deeply.
5.5. Synthesis
Overall, the discussion underscores how technological readiness alone is insufficient for the societal integration of autonomous public transport. Adoption hinges on the alignment of system design and communication with the psychological diversity of users. The challenge for planners and policymakers lies in bridging these motivational divides, thereby transforming automation from a technological innovation into a socially legitimate and accepted mobility option.
6. Conclusion
This study investigated how different groups within the Munich Metropolitan Region vary in their mobility preferences and potential acceptance of autonomous public transport. Using a combination of SEU modeling and cluster analysis, five user types were identified:
Environmentalists, Pragmatists, Indifferent Moderates, Comfort-Oriented Users and Cost- and Environmentally Conscious Users.
These clusters represent distinct motivational structures – ranging from eco-rational to performance-rational orientations – that explain diverging attitudes toward autonomous buses. Clusters 1 and 5 are potential early adopters driven by sustainability and affordability. Clusters 2 and 4 show resistance rooted in autonomy, comfort, and control needs. Cluster 3 occupies a neutral middle ground, likely to follow once social normalization occurs.
The results extend behavioral transport segmentation by integrating psychographic clustering with utility-based modeling, demonstrating that acceptance of automation is shaped by both value systems and instrumental evaluations. For policy and planning, this implies that uniform communication and service strategies will fail. Instead, system design and outreach must reflect motivational diversity: This must emphasize ecological and social benefits for sustainability-driven users, highlight performance, reliability, and convenience for efficiency- and comfort-oriented users, and strengthen social visibility and trust to convert neutral users into regular passengers.
Pilot projects should include participants from all clusters, ensuring representative feedback and equitable system design. While the study provides a robust snapshot of mobility typologies, it captures stated intentions rather than observed behavior and may overrepresent digitally engaged participants. Ultimately, the integration of autonomous buses will depend not only on technical feasibility but on human alignment: the extent to which these systems resonate with people’s values, expectations, and everyday mobility cultures. Recognizing this diversity is key to transforming automation from a technological innovation into a socially accepted and sustainable mode of urban transport.
Acknowledgement
Open access
