Negotiating Autonomy: A Structured Literature Review of Equity and Governance Dimensions Within Autonomous Vehicle Acceptance Research

Abstract

Autonomous vehicle technology has rapidly advanced in recent years. Such technology is increasingly viewed not merely as a technical innovation but also as a social and behavioural transformation shaped by how these systems are interpreted, trusted, and integrated into everyday life. There are mounting expectations regarding its potential to improve traffic safety, enhance energy efficiency, reduce congestion, and support sustainable mobility; however, key questions remain about how different groups and communities experience autonomous mobility. This review synthesizes equity, governance, and sustainability dimensions as they appear within the existing corpus of AV user acceptance research. A structured review of research on autonomous vehicles (AVs) and user acceptance was conducted using an initial database search followed by iterative literature refinement and structured thematic coding. Using this approach, the review identifies key thematic patterns, highlights structural research gaps, and explores regional differences, offering a framework that supports subsequent comparative analysis. AVs have the potential to shape accessibility, social relations, and sustainable lifestyles. By integrating technological advancement with local governance, community practices, and social equity considerations, automated public transit may serve as a catalyst for sustainable and inclusive urban transformation.

1. Introduction

1.1. Evolution of AVs Development

Autonomous vehicle technology has evolved rapidly in recent years, accompanied by the growing expectations for its potential to improve traffic safety, enhance energy efficiency, reduce congestion, and support sustainable mobility [1,2,3,4,5,6,7]. As core algorithms, sensor systems, and vehicle-infrastructure coordination technologies continue to mature [8,9], research has also addressed network-level optimization challenges, including Pareto-optimal path generation under stochastic conditions [10] and reliable vehicle routing using traffic sensor–augmented information [11]. Some scholars have noted a shift in Autonomous Vehicles (AVs) applications from private vehicle settings toward shared mobility and autonomous public transportation services [12,13]. At the same time, researchers increasingly emphasize that autonomous driving is not solely a technical breakthrough but also a social and behavioural transformation, shaped by how individuals interpret, trust, and integrate these systems into everyday life [14,15,16]. Accordingly, public acceptance has become a central issue in global discussions on the deployment of AVs [17].

Research on public acceptance of AVs shows a clear shift since 2020. Early studies in the 2000s focused mainly on technical development, including system performance, algorithms, and safety [18,19,20]. From around 2010 onward, scholars began to examine how the public perceives and accepts autonomous vehicle technologies. After 2020, the volume of acceptance-related research increased sharply and has continued to grow. The discussion also became more diverse and detailed, covering different user groups, travel contexts, and social factors. Recent studies place greater emphasis on the attitudes and behavioural responses of passengers, drivers, and other road users [21,22,23,24,25]. One reason for this shift is that autonomous vehicle technologies have matured, enabling real-world pilots and trials in more cities and localized settings, providing researchers with more realistic data and observations [26]. In addition, the rise in contactless mobility needs during the COVID-19 pandemic drew greater commercial and policy attention to AVs, further accelerating research activity in this area [27,28,29]. Despite this rapid growth, many key questions about how different groups and communities experience autonomous mobility remain insufficiently examined.

1.2. Gaps in Existing Research on AV Acceptance

While large studies collectively provide a strong analytical foundation, the literature simultaneously highlights several ongoing gaps. First, as shown by Pande and Taeihagh [30], Saravanos, Pissadaki [31], Zhou, Guo [32], Luo, He [33], the majority of existing research and investigation examines acceptance at the national or city level, with relatively limited attention to the community scale, where everyday mobility experiences are embedded [34]. Second, cross-cultural comparative studies remain scarce, and questions persist regarding whether constructs such as trust and acceptance operate equivalently across different cultural contexts [4,35,36]. Third, scholars have noted that vulnerable groups are often underrepresented in empirical samples, including older adults, children, persons with disabilities, and non-motorized road users [37,38]. Fourthly, as argued by Zarbakhshnia and Ma [16] and Grindsted, Christensen [39], current research rarely connects AV acceptance to broader Sustainable Development Goals (SDGs), resulting in insufficient attention to issues of social equity and inclusion, and the need for community and climate resilience.

These gaps are particularly evident in the emerging field of autonomous public transportation, where real-world deployments have begun to outpace systematic academic analysis. The use of AVs in public transportation systems has received increasing attention in recent years, as highlighted by Paddeu, Shergold [40], Carrese, Sportiello [12], Gusev and Gilroy [41] and Ross [42]. While pilot projects across various regions have demonstrated the technological feasibility of such systems, and more studies in recent years have begun to examine public acceptance and attitudes, systematic research addressing accessibility, fairness, and the experiences of vulnerable groups in real operational scenarios remains relatively limited, as noted by Schandl, Fischer [43] and Golbabaei, Dwyer [44]. Much of the existing research continues to be confined to short-term test trips or simulated environments, which, as pointed out by Debbaghi, Rombaut [45], Derakhshan, Nosrat Nezami [46], Li, Shi [47], and Sun, Huang [48], cannot fully capture the diverse needs and travel patterns of residents in their daily lives. Furthermore, in-depth analyses of accessibility, equity, and community disparities of public AV services remain limited, as discussed by Schiller and Kenworthy [49]. These gaps suggest that understanding AV acceptance requires shifting attention from national or city-level patterns to where mobility is lived, at the community level.

1.3. Importance of Community-Level Mobility Patterns

In many communities, particularly those with ageing populations, areas experiencing transport-poor conditions, and ‘self-contained’ urban neighborhoods (i.e., urban neighborhoods where walking, cycling, and local transit are predominant), residents’ mobility needs are highly localized [50,51]. These characteristics mean that short-distance travel, access to services, and social participation vary significantly across communities. Overlooking these differences may exacerbate existing inequalities rather than alleviate them, as noted by Bokolo [52]. AV acceptance and equity need to account for these differences rather than assuming uniform conditions across the city. Against this backdrop, the community scale provides an essential lens for understanding the societal impacts of AVs, as highlighted by Schiller and Kenworthy [49]. Most daily travel occurs over short distances, which, according to Hryniewicz and Grzegorczyk [53] and Yalciner Ercoskun, Kajosaari [54], form the foundation for maintaining social connections and resilience. For populations such as older adults, children, and low-income households, transport poverty directly affects their quality of life and ability to participate socially, as emphasized by Bokolo [52]. Yet, despite these insights, existing research seldom investigates how autonomous vehicle technologies interact with community-level practices, cultural norms, and day-to-day routines. As suggested by Haapio [50], a community-oriented perspective, therefore, not only helps to fill these knowledge gaps, but also opens new avenues for examining how equity, public acceptance, and sustainable mobility converge in the context of autonomous transportation systems.

Although several review studies have examined autonomous mobility, many have primarily concentrated on technological performance, safety modelling, or bibliometric trend mapping [55,56,57]. While these contributions provide valuable overviews of technical development and publication trajectories, fewer reviews synthesize how governance, equity, and cross-cultural variation are treated within AV user acceptance research. As a result, the social and contextual dimensions of AV acceptance remain fragmented across the literature. This review seeks to address that gap by offering a structured thematic synthesis grounded in social science perspectives.

Building on these observations, this structured literature review seeks to retrace how research on AV acceptance has evolved over the past decade, paying particular attention to how issues of equity, cross-cultural variation, vulnerable groups, and community context have been incorporated or neglected in existing studies. By synthesizing evidence across different regions and methodological traditions, the review aims to re-examine the role of autonomous mobility in sustainable urban development and to highlight its implications for future policy and research. Beyond considering how the public evaluates autonomous mobility, this review also asks which groups are underrepresented, which contexts remain insufficiently explored, and how autonomous mobility is interpreted and implemented across different cultural and community settings. Such a perspective enables a more comprehensive understanding of the societal impacts of AVs. It supports their development in ways that are more inclusive, equitable, and conducive to long-term sustainability.

To guide the analysis, this review addresses three research questions:

RQ1: How has existing AV user acceptance research conceptualized and examined equity considerations?

RQ2: What governance, cultural, and contextual factors are identified as influencing adoption and implementation patterns within the AV acceptance literature?

RQ3: What conceptual and empirical gaps remain underexplored, particularly at community and cross-cultural levels?

This study does not aim to replicate a PRISMA-based systematic review or bibliometric analysis. Instead, it adopts a structured thematic literature review approach to synthesize social science perspectives on autonomous mobility. Its contribution lies in integrating thematic coding, cross-cultural comparison, and governance-oriented analysis to identify patterns of concentration and underexplored areas within the literature, particularly in relation to equity and medium-sized urban contexts.

2. Conceptual & Theoretical Background

2.1. Technological and Functional Landscape of AVs

AVs are typically classified into Levels 0–5 based on their level of automation. This classification system was proposed by the Society of Automotive Engineers (SAE) and is systematically defined in its SAE J3016 standard [58]. Each level reflects the degree to which the vehicle replaces human drivers in dynamic driving tasks, ranging from full human control (Level 0) to full automation (Level 5). This review organizes the definitions for each automation level according to the SAE J3016 standard, as shown in

Table 1. SAE J3016 Levels of Driving Automation.

Level	Automation Type	Vehicle Role	Driver Role/Responsibility	Example Applications
0	No Automation	None	Driver performs all driving tasks	Standard manual cars
1	Driver Assistance	Assistance for specific functions (steering or acceleration/braking)	Driver monitors environment and performs most driving tasks	Adaptive cruise control, lane keeping assist
2	Partial Automation	Vehicle controls both steering and acceleration/braking simultaneously	Driver monitors environment, ready to intervene at any time	Tesla Autopilot (partial features), GM Super Cruise
3	Conditional Automation	Vehicle handles all driving tasks under specific conditions	Driver must be available to intervene when requested	Highway pilot under limited conditions
4	High Automation	Vehicle performs all driving tasks within a specific operational design domain (ODD)	Driver may not need to intervene in ODD	Autonomous shuttles in defined urban zones, closed-campus mobility
5	Full Automation	Vehicle handles all driving tasks under all conditions	Driver not required	Fully autonomous taxis, robotaxis in all environments

Source: Adapted from SAE J3016, Society of Automotive Engineers [58].

. The evolution from partial to full automation reflects not only increasing technical complexity but also broader societal and systemic implications. Currently, most AVs on the market remain at Levels 2–3, falling within the semi-automated category. Full automation has yet to become widespread, though related technologies are advancing rapidly, as noted by Choi, Min [13]. As core sensing, perception, and control technologies have gradually matured, large-scale deployment has increasingly become a key objective for the industry, particularly as a means of reducing costs and improving operational feasibility [59].

In terms of vehicle types, autonomous driving can be categorized into private autonomous vehicles, shared autonomous vehicles (SAVs), and public autonomous transportation systems, a distinction drawn from studies such as those by Silvestri, De Fabiis [23] and Soltani, Ananda [60]. These classifications also align with the Tech Codes system used in this review. Different technical models target distinct usage scenarios and user groups, potentially leading to varying levels of acceptance and equity outcomes, as discussed by Tian and Wang [61]. For instance, Chng, Anowar [62] note that private AVs are more likely to serve specific income brackets, whereas shared or public autonomous systems tend to be closely associated with urban accessibility, equity, and sustainability goals. Understanding these distinctions is crucial for further analyzing societal acceptance and policy pathways.

AVs can be regarded as a key technology for advancing sustainable transportation, with studies such as Adloff [63] pointing to potential environmental benefits, including reduced emissions, alleviated traffic congestion, and improved energy efficiency. They may also help operators lower costs and extend transit services in both urban and rural areas through automated public transport, as indicated by Gusev and Gilroy [41]. However, scholars, including Fielbaum [34] emphasize that these positive impacts depend on AVs being deployed thoughtfully and integrated effectively with existing transport systems. Improper deployment may instead intensify mobility challenges, such as increased empty-mileage travel, congestion, urban sprawl, and more complex parking demands, concerns raised by Ariza-Álvarez, López-Lambas [64]. In addition, public perceptions of AVs extend beyond notions of convenience, such as hands-free operation. Safety, data privacy, and anxieties about automation replacing human jobs all remain important factors shaping societal acceptance.

2.2. Individual-Level Acceptance and Behavioural Foundations

2.2.1. Theoretical Frameworks and the Role of Trust

In studies of new technology adoption, scholars have identified the Technology Acceptance Model (TAM) (TAM (Technology Acceptance Model): Focuses on perceived usefulness and perceived ease of use as primary determinants of technology) developed by Davis [65] and the Theory of Planned Behaviours (TPB) (Theory of Planned Behaviours (TPB), which emphasizes attitudes, subjective norms, and perceived behavioural control as determinants of intention) proposed by Ajzen [66] as the earliest widely cited theoretical frameworks. Subsequently, Venkatesh, Morris [67] introduced the Unified Theory of Acceptance and Use of Technology (UTAUT), Unified Theory of Acceptance and Use of Technology (UTAUT), which integrates multiple acceptance models and identifies performance expectancy, effort expectancy, social influence, and facilitating conditions as core predictors, which has since been extended by some researchers to UTAUT2 (UTAUT2, an extension of UTAUT that adds hedonic motivation, price value, and habit for consumer contexts) in recent years [68]. A summary of these frameworks, their core factors, and their applications in AV research is provided in Table 2. While TAM and UTAUT remain the most frequently adopted frameworks in studies of autonomous driving technology acceptance, other models, including TPB, continue to be applied to examine behavioural intentions and adoption patterns in specific settings. These frameworks emphasize that performance expectancy, perceived ease of use, and social influence are central factors shaping individuals’ willingness to adopt new technology. A substantial body of empirical work has since adopted frameworks such as TAM and UTAUT to investigate determinants of acceptance, including trust, perceived safety, perceived usefulness, and social influence [2,4,6,68,69,70,71,72].

With the rapid advancement of autonomous vehicle technology, Koskinen, Mallat [73], Kaur and Rampersad [74], and Dirsehan and Can [75] have highlighted that trust has gradually emerged as a crucial factor influencing user acceptance. Further research by Paddeu, Parkhurst [76] and Golbabaei, Dwyer [44] subdivides trust into distinct dimensions, including performance trust, institutional trust, manufacturer trust, and trust in system reliability and transparency, providing a more nuanced explanation of how users evaluate complex intelligent systems.

Table 2. Comparison of Key Technology Acceptance Models Applied in AV Research.

Model	Core Factors	Application in AVs Studies	Key References
TAM	Perceived usefulness, Perceived ease of use	User acceptance and behavioural intention	Davis [65]
TPB	Attitude, Subjective norms, Perceived behavioural control	Behavioural intention in transportation context	Ajzen [66]
UTAUT	Performance expectancy, Effort expectancy, Social influence, Facilitating conditions	AV adoption across user groups	Venkatesh et al. [67]
UTAUT2	UTAUT factors and Hedonic motivation, Price value, Habit	Consumer adoption contexts in AVs	Venkatesh et al. [77], Korkmaz et al. [68]

Source: Adapted from Davis [65], Ajzen [66], Venkatesh et al. [67,77], and Korkmaz et al. [68].

Table 2. Comparison of Key Technology Acceptance Models Applied in AV Research.

Model	Core Factors	Application in AVs Studies	Key References
TAM	Perceived usefulness, Perceived ease of use	User acceptance and behavioural intention	Davis [65]
TPB	Attitude, Subjective norms, Perceived behavioural control	Behavioural intention in transportation context	Ajzen [66]
UTAUT	Performance expectancy, Effort expectancy, Social influence, Facilitating conditions	AV adoption across user groups	Venkatesh et al. [67]
UTAUT2	UTAUT factors and Hedonic motivation, Price value, Habit	Consumer adoption contexts in AVs	Venkatesh et al. [77], Korkmaz et al. [68]

Source: Adapted from Davis [65], Ajzen [66], Venkatesh et al. [67,77], and Korkmaz et al. [68].

2.2.2. Public Attitudes and Broader Social Influences

However, as Paddeu, Parkhurst [76] point out, these traditional acceptance frameworks have significant limitations. Most models primarily focus on individual-level psychological and behavioural intent, while overlooking broader structural and societal factors, such as fairness, cultural differences, the needs of socially vulnerable groups, community contexts, and Sustainable Development Goals (SDGs) [78]. In the context of autonomous vehicles, Wicki, Guidon [79] emphasize that user acceptance is influenced not only by personal attitudes and trust but also by city-scale characteristics, travel patterns, the distribution of community resources, the needs of vulnerable road users, and the cultural interpretation of risk and technology. Hadid, Irawan [80] and Ganga, Avanzi [70] similarly note that public attitudes are shaped not only by individual experiences and perceptions of system characteristics but also by broader social and cultural factors. To provide a clearer overview, Table 3 summarizes the key social and community-level factors identified in the literature on public acceptance of autonomous vehicles.

Experts argue that the successful deployment of autonomous vehicles should not be treated solely as a technical optimization of transportation systems. Instead, the core challenge lies in promoting AV adoption across diverse urban and cultural contexts while aligning with local residents’ real mobility needs. Ensuring that technological advancement does not exacerbate mobility burdens for vulnerable groups but rather enhances the accessibility and inclusivity of the overall transport system is critical. Only when autonomous vehicles are designed and implemented with attention to everyday mobility practices, local community characteristics, and diverse user needs, and when these systems are integrated carefully with existing transport networks, can they meaningfully contribute to the long-term sustainable development of cities. Among the various factors influencing public attitudes, Hadid, Irawan [80] highlight safety perception and risk cognition as the most frequently discussed. This includes judgments on vehicle safety, system stability, cybersecurity, collision prevention capabilities, and data privacy risks. Jing, Wang [81] further note that media coverage, public incidents, and societal narratives play a major role in shaping public understanding, where positive narratives may raise expectations for the technology, while accidents or negative events can amplify concerns, as Quinones, Romero [82] observe.

Notably, Long, Shi [21] emphasize that most studies focus on passengers’ attitudes, even though autonomous shuttles must share roads with other users such as pedestrians and cyclists. Creutzig, Javaid [83] argue that AVs need to account for the safety and needs of these non-motorized users, given their equal status as road users but relative lack of dedicated infrastructure. Furthermore, Nordhoff, Kyriakidis [84] and Portouli, Karaseitanidis [85] demonstrate that real-world pilot projects, including test rides, on-site observations, and participatory evaluations, can substantially influence public perception and trust, helping residents gain a concrete understanding of vehicle behaviour and system reliability. Taken together, Wu, Wang [86] suggest that public attitudes can be regarded as an extension of traditional acceptance models such as TAM and UTAUT, offering insight into the broader social dynamics of AV acceptance. While Wai Mun and Lee [69] note that the public generally acknowledges the potential of autonomous vehicles to reduce driving burdens, they also remain concerned about technological reliability, personal privacy protection, and integration with existing transportation systems. These considerations collectively constitute the key mechanisms shaping societal acceptance of autonomous vehicles.

Table 3. Social and Community-Level Factors Influencing Public Acceptance of AVs.

Factor Category	Key Dimensions	Description	Key References
Individual perceptions	Safety perception, trust, risk cognition	Perceptions of vehicle safety, system reliability, cybersecurity, and data privacy	Hadid et al. [80]; Jing et al. [81]
Social and cultural context	Cultural norms, risk interpretation	Cultural values shape how risk and automation are understood	Wicki et al. [79]; Ganga et al. [70]
Community characteristics	Local mobility patterns, resource distribution	Acceptance is influenced by daily travel routines and access to services	Wicki et al. [79]; Haapio [50]
Vulnerable road users	Pedestrians, cyclists, older adults	Non-motorized and vulnerable users face distinct safety and access concerns	Long et al. [21]; Creutzig et al. [83]
Media and public discourse	Media coverage, incidents, narratives	Media framing and public events shape expectations and fears	Jing et al. [81]; Quinones et al. [82]
Experiential exposure	Pilot projects, test rides	Direct experience increases understanding and trust	Nordhoff et al. [84]; Portouli et al. [85].

Note: References listed are illustrative rather than exhaustive.

Table 3. Social and Community-Level Factors Influencing Public Acceptance of AVs.

Factor Category	Key Dimensions	Description	Key References
Individual perceptions	Safety perception, trust, risk cognition	Perceptions of vehicle safety, system reliability, cybersecurity, and data privacy	Hadid et al. [80]; Jing et al. [81]
Social and cultural context	Cultural norms, risk interpretation	Cultural values shape how risk and automation are understood	Wicki et al. [79]; Ganga et al. [70]
Community characteristics	Local mobility patterns, resource distribution	Acceptance is influenced by daily travel routines and access to services	Wicki et al. [79]; Haapio [50]
Vulnerable road users	Pedestrians, cyclists, older adults	Non-motorized and vulnerable users face distinct safety and access concerns	Long et al. [21]; Creutzig et al. [83]
Media and public discourse	Media coverage, incidents, narratives	Media framing and public events shape expectations and fears	Jing et al. [81]; Quinones et al. [82]
Experiential exposure	Pilot projects, test rides	Direct experience increases understanding and trust	Nordhoff et al. [84]; Portouli et al. [85].

Note: References listed are illustrative rather than exhaustive.

2.3. Societal, Spatial, and Contextual Dimensions

2.3.1. Equity and Vulnerable Groups

Scholars have noted that urban expansion, road construction, and increased private vehicle use are reshaping residents’ travel patterns while simultaneously undermining low-carbon mobility options [87]. In response, some researchers argue that the development of AV technology must align with existing local sustainable transportation practices to avoid reinforcing structural pressures in mobility systems [12].

Research on transportation equity emphasizes that children, the elderly, people with disabilities, low-income individuals, pedestrians, and cyclists should all have equal access to mobility opportunities [37,38,88]. Scholars warn that, although autonomous vehicle technology has the potential to enhance mobility for these groups, uneven deployment, high cost barriers, or biased system designs could generate new forms of inequality [89]. Fielbaum [34] further observes that underrepresentation of vulnerable groups in pilot projects often hinders their needs from being incorporated into system improvements.

Regarding technology acceptance, numerous studies have employed models such as TAM and UTAUT to investigate public attitudes [68,86,90,91]. However, scholars point out that these studies predominantly focus on national or city-level scales, while community-level short trips, including walking, cycling, and local shuttle use, receive limited attention, despite being particularly important for seniors and children [52,92]. Moreover, Madigan, Louw [93] highlight that research on diverse user groups such as pedestrians, cyclists, non-autonomous vehicle drivers, and individuals with disabilities remains relatively scarce. Collectively, these studies indicate that equity and inclusivity are central considerations for the deployment of autonomous vehicles.

2.3.2. Impacts on Urban Sustainability

In recent years, scholars and practitioners have begun to explore the use of AVs in areas related to sustainability, including logistics optimization, environmental monitoring, and even wildlife conservation, where automated vehicles can support low-impact access, data collection, and habitat protection [94]. Wai Mun and Lee [69] suggest that autonomous driving offers potential benefits for urban sustainability, including emissions reduction, optimized energy usage, and enhanced traffic efficiency. Yet, other researchers caution that unintended consequences may arise: for example, increased empty-mileage driving could exacerbate congestion, longer commutes might accelerate urban sprawl, and private autonomous vehicles could displace public transport, reducing overall system efficiency. Korkmaz, Fidanoglu [68] argue that shared or public autonomous vehicles may offer greater environmental benefits, although their effectiveness depends on factors such as urban spatial structure, operational models, and passenger occupancy rates.

From a systemic perspective, scholars highlight that the development of transport automation is influenced by path dependency, system lock-in, and pre-existing travel habits [95]. Schreurs and Steuwer [96] note that, although European studies often link autonomous vehicles with sustainability discussions, the technology has not yet been fully integrated into city-level sustainable transport strategies. In addition, Dirsehan and Can [75] emphasize that evaluations of new transport technologies should consider multidimensional factors, including acceptability, affordability, accessibility, availability, and adaptability. Hinderer, Stegmüller [97] warn that if vulnerable groups are unable or unwilling to adopt new technologies, social exclusion may increase. Meanwhile, González-González, Nogués [98] discuss governance challenges, noting that questions about liability attribution and certification mechanisms for autonomous vehicles reveal gaps in current policy frameworks. Collectively, these perspectives suggest that sustainable autonomous mobility requires coordination among technological innovation, institutional design, governance, and social acceptance, ensuring that new systems are adopted and integrated fairly within communities.

2.3.3. Cultural and Community Contexts

Scholars have observed that cultural factors play a significant role in shaping public attitudes toward autonomous vehicles. Research drawing on Hofstede’s cultural dimensions indicates that risk aversion, power distance, collectivism, and uncertainty tolerance influence public risk perception, institutional trust, and expectations regarding automated systems [43,99]. Social-technical transformation theory further highlights that technology adoption is not determined solely by technical performance but is shaped by the interplay of institutional structures, governance logics, cultural norms, and everyday social practices [54]. Within cities, scholars argue that community-level characteristics, including demographic composition, commuting patterns, road safety experiences, infrastructure conditions, and public transit accessibility, significantly influence residents’ trust, demand, and willingness to adopt autonomous services [53]. For instance, Berrada, Mouhoubi [100] point out that ageing communities or transport poverty neighbourhoods often exhibit heightened safety sensitivities and distinct mobility needs, resulting in localized evaluation frameworks for autonomous services. Conversely, in communities with more developed transportation systems or higher digital literacy, autonomous vehicles tend to be perceived as a complementary service rather than a replacement. Haapio [50] and Taniguchi, Enoch [101] argue that integrating cultural differences with community contexts helps explain variations in acceptance across national, urban, and neighbourhood scales, offering critical insights for context-sensitive deployment strategies.

3. Methodology

3.1. Search & Selection

This study adopts a structured literature review approach to identify and categorize thematic patterns within social science research on autonomous mobility and user acceptance. The objective of the review is to synthesize conceptual developments and identify research gaps rather than to conduct exhaustive enumeration or quantitative meta-analysis. The full list of included studies is publicly available in Appendix A: Literature Database to enhance transparency and reproducibility. The literature library will also be made accessible via the Open Science Framework (OSF) upon finalization.

An initial structured search was conducted in the Scopus database using the keyword combinations ‘Autonomous Vehicle’ AND ‘User Acceptance’ and ‘Autonomous Shuttle Bus’ AND ‘User Acceptance’. The search was limited to journal articles published in English (English-language publications were selected for three main reasons: first, English has become the primary language of academic communication, with most high-impact journals publishing their latest findings in English; second, English-language literature offers more systematic database coverage, peer-review standardization, and citation traceability; third, restricting the review to a single language enhances replicability and reduces potential bias arising from cross-linguistic interpretation) between 2015 and 2025 and restricted to subject areas including social sciences, psychology, and multidisciplinary studies to ensure relevance to behavioral and governance dimensions. The search covered publications from January 2015 to November 2025. Articles indexed as 2026 in Scopus but available online ahead of print during the search period were also included. The final database search was completed on 25 November 2025.

While terminology in this field continues to evolve (e.g., automated vehicles, self-driving systems, shared autonomous vehicles), the consistent use of “user acceptance” as a core construct was retained to maintain conceptual clarity and comparability across studies. This focus inevitably excludes some adjacent adoption-oriented research and is acknowledged as a boundary of the review.

The Scopus searches yielded 218 records. Titles and abstracts were manually screened by the authors, and 15 non-land-transportation studies (e.g., drones, autonomous vessels) were excluded, resulting in 203 articles retained for further analysis. After the structured Scopus search and screening process (203 retained articles), a second identification stage was conducted through citation-based identification and iterative snowballing during full-text review. This process aimed to capture conceptually relevant literature embedded within the AV user acceptance research network that may not have been retrieved through keyword-based database indexing alone. Unlike the core Scopus search, this stage did not constitute a parallel systematic database retrieval with predefined query strings. Instead, references cited within included articles were screened against the same predefined eligibility criteria. Studies were included if they met the temporal scope (2015–2025), focused on autonomous land-transport systems, and addressed user acceptance or related social dimensions (e.g., equity, governance, culture, sustainability). As this process was citation-driven and iterative, it is not possible to retrospectively attribute each supplementary record to a specific database source. However, all inclusion decisions were applied consistently based on the predefined criteria. Records from both stages were cross-checked for potential duplication using DOI and title matching within reference management software prior to final consolidation. No duplicate records were identified across the combined corpus. The complete corpus of 284 studies is provided in Appendix A to ensure transparency of the final dataset. In total, 81 additional studies were incorporated through this citation-based enrichment stage. A summary of the identification and screening stages is provided in Appendix B.

Inclusion and coding decisions were conducted iteratively by the first author and reviewed by the second author at multiple stages. Rather than applying formal inter-rater reliability testing, a structured audit process was undertaken in which a subset of coded studies was examined to assess thematic alignment and conceptual boundary clarity. Ambiguities were documented and used to refine coding definitions prior to final synthesis. Ambiguities were documented and used to refine coding definitions prior to final synthesis. Given the objective of mapping thematic orientations rather than evaluating causal validity or effect sizes, formal risk-of-bias assessment tools were not employed. Instead, methodological type was recorded to allow readers to interpret thematic distributions in light of study design variation. Interpretations regarding research gaps therefore refer to patterns of thematic attention rather than normative judgments of study quality. The final corpus reflects a conceptually curated body of literature suitable for thematic synthesis and comparative analysis.

3.2. Coding & Synthesis

Following study inclusion, all articles were subjected to a structured coding framework and thematic synthesis to analyze patterns across the literature. Articles were coded along four analytical dimensions: research theme, methodology, regional context, and technology type. This framework was designed to capture research priorities and topic distributions, identify methodological trends, compare regional patterns in AV adoption, and distinguish the influence of different technological scenarios on reported findings. Data items extracted from each study included research focus, methodological approach, regional setting, technology type, and key findings related to user acceptance, equity, governance, cultural context, and sustainability. The full coding scheme and definitions are summarised in

Table 4. Coding framework for the structured literature review.

Coding Dimension	Code	Category Description
Research Theme	A	User Acceptance & Public Perception
	B	Equity, Accessibility & Inclusion
	C	Governance & Policy Frameworks
	D	Public Participation & Engagement
	E	Sustainability & System Integration
	F	Urban & Cross-cultural Contexts
Methodology	QNT	Quantitative
	QLT	Qualitative
	MIX	Mixed-methods
	REV	Review paper
	THEO	Theoretical/Conceptual
	SIM	Simulation/Modelling
Technology Type	AV	Autonomous Vehicle
	SAV-S	Shared-use SAV (multi-occupancy)
	SAV-P	Private-use SAV
	BOTH	Both shared and private
	NA	Not specified
Region	EUR	Europe
	NAM	North America
	ASI	Asia
	LAC	Latin America & Caribbean
	MENA	Middle East & North Africa
	OCE	Oceania

Note: To ensure analytical clarity and comparability, each article was assigned to a single primary code per category based on its main research emphasis.

.

To enhance transparency and reproducibility, the complete list of included studies is available in the Appendix A: Literature Database, which is also publicly accessible via OSF. Table 4 presents an overview of the coding framework and the distribution of studies across research themes, methodologies, regions, and technology types.

In addition to categorical coding, this review adopts a structured keyword-based synthesis to support the thematic analysis presented in the Results section. For each research theme, keywords related to four analytical dimensions: Safety and Trust, Equity, Governance, and Culture, were identified based on recurring concepts in the reviewed literature and prior theoretical discussions. Keyword occurrences were examined within article abstracts to capture patterns of emphasis across themes. Each article was assessed for the presence of predefined keywords corresponding to the four analytical dimensions (Safety & Trust, Equity, Governance, and Culture). Shading intensity in the matrix reflects the relative emphasis of each dimension within a theme, derived from this keyword-assisted abstraction and qualitative synthesis, rather than precise quantitative counts. The keyword-based abstraction served as a heuristic aid to thematic interpretation rather than as a standalone text-mining procedure. This approach supports comparative interpretation, highlights general patterns of concentration and underrepresentation, and provides a structured basis for subsequent analysis of thematic differences across regions.

Primary vs. Secondary Topic Assignment: For each article, a primary research theme was assigned based on the main focus as identified in the abstract, methodology, and key findings. Secondary topics were noted when the study addressed additional themes but were not the central emphasis. This distinction ensures that each article contributes primarily to one thematic category while still acknowledging the presence of related concepts across other themes. All assignments were discussed among the authors to ensure interpretive consistency.

4. Results

4.1. Thematic Distribution and Core Findings

Figure 1 shows that existing research on AV user acceptance is highly concentrated on Theme A (User Acceptance and Safety Trust), which constitutes the bulk of the literature. Themes B (Equity and Accessibility) and Theme E (Sustainability) make up much smaller shares of the literature, while Governance (Theme C), Public Participation (Theme D), and Cross-Cultural and Urban Contexts (Theme F) remain largely marginal. This pattern shows that current research still focuses on individual-level technology acceptance and risk perception, with only limited attention to institutional, social structural, and spatial variations.

Figure 2 further reveals differences in focus within each theme. Safety and trust occupy a central position across all themes, particularly in Theme A (User Acceptance and Safety Trust), Theme C (Governance), and Theme F (Cross-Cultural and Urban Contexts). The equity dimension is primarily concentrated in themes B (Equity and Accessibility) and E (Sustainability), while the cultural dimension remains generally weak, with only theme F (Cross-Cultural and Urban Contexts) showing a clear focus. This distribution indicates that although ‘urban and cross-cultural contexts’ is explicitly defined as an independent theme, related research often discusses cultural differences indirectly through safety or governance perspectives rather than treating them as core analytical subjects.

4.2. Methodological Distribution of AV User Acceptance Research

Figure 3 shows that most AV user acceptance studies use quantitative methods. There are far fewer papers using qualitative, mixed-methods, or theoretical approaches. This indicates that the field remains centered on model validation, variable relationships, and attitude measurement as core research pathways. In contrast, qualitative research and theoretical/conceptual studies are limited in number, while simulation and modelling research occupy a distinctly marginal position.

Figure 4 further reveals methodological variations within different research themes. Looking at the themes, A (User Acceptance and Safety Trust) and F (Cross-Cultural and Urban Contexts) mostly use surveys and stats. Themes D (Public Participation) and C (Governance) have more interviews, case studies, or policy discussions. Themes B (Equity and Accessibility) and E (Sustainability) use a mix, but still, there are not many studies, and most are small or limited in scope. Collectively, this methodological structure reveals a structural tension: research subjects are increasingly complex (involving institutions, equity, and social practices), yet dominant methods remain centred on individual-level quantitative measurement. This partially limits the in-depth understanding of long-term, contextual, and community-level issues.

4.3. Technological Focus of AV Acceptance Studies

Figure 5 indicates that existing research is highly concentrated on private AVs, with SAVs, particularly in the ‘shared but privately ridden’ form, being notably underrepresented in the literature. This distribution reflects that academic discourse remains primarily centered on individual mobility scenarios, while attention to systemic, collective mobility patterns remains relatively limited.

Figure 6 reveals differences in the emphasis on technology types across research themes. Themes A (User Acceptance and Safety Trust), C (Governance), D (Public Participation), and F (Cross-Cultural and Urban Contexts) primarily address AVs, focusing on safety, governance, and user acceptance in cities. Themes B (Equity) and E (Sustainability) include more shared, multi-passenger SAVs, since these are often linked to social equity and environmental goals. Still, SAVs are studied less than AVs overall, and AVs remain the focus in most research themes. This technological distribution reveals an inherent tension in current research: despite SAVs often being assigned higher expectations for equity and sustainability, the scale of empirical research has yet to match their policy significance.

4.4. Regional Distribution of AV Acceptance Research

As illustrated in Figure 7, most studies are drawn from Europe, Asia, and North America, with a smaller but visible group adopting global or comparative approaches. Research centered on Latin America, the Middle East and North Africa, and Oceania is much less common, suggesting that these regions have received limited empirical attention so far. When examined by theme, geographical patterns become more uneven (Figure 8). Research under Themes A (User Acceptance and Safety Trust) and D (Public Participation) is largely anchored in European contexts, while Themes C (Governance) and F (Cross-Cultural and Urban Contexts) more often rely on multi-country or comparative approaches. Even so, studies labelled as ‘urban and cross-cultural’ rarely engage with detailed cases from non-Western or small- and medium-sized cities. Instead, it relies more on macro or transnational comparisons to present cultural differences. Overall, this geographical structure indicates that while cross-cultural discussions are not formally absent, empirical depth at the level of specific cities, communities, and local contexts remains limited.

4.5. Synthesis: What Is Debated and What Is Missing

Figure 9 draws together the key points of contention and the more persistent gaps that emerge from the preceding analysis of themes, methods, technologies, and regions in the AV user acceptance literature.

In terms of debate, the literature repeatedly returns to a small set of questions. One concerns whether AVs are likely to ease urban congestion or, conversely, generate additional traffic. Other centers on whether automation can meaningfully contribute to social equity. A third focuses on the sustainability trade-offs between private AVs and SAVs. These debates are predominantly distributed across themes A, B, and E, reflecting discussions centred on system performance, user attitudes, and macro-level objectives.

In contrast, several recurring areas appear comparatively less developed in the existing literature. These include community-level embedding processes, long-term longitudinal empirical data, the role of cultural differences in governance and implementation, and the integration of AVs into everyday mobility practices. These themes (particularly F and C) receive comparatively limited empirical and methodological attention. This pattern suggests that current research on AV user acceptance is more concentrated on normative and attitudinal debates, while empirical investigations of how acceptance unfolds within specific cities and communities remain relatively limited.

5. Discussion

5.1. Interpretation of Key Findings

This review reveals that the literature on AV user acceptance exhibits a markedly imbalanced structure across themes, methodologies, technological subjects, and regional coverage. Existing research is heavily concentrated on safety, trust, and individual-level technology adoption, while topics such as fairness, governance, culture, and community contexts remain consistently marginalized. This has fostered a research landscape centered on ‘acceptability’ with ‘social consequences’ relegated to secondary importance. These themes overlap rather than being completely separate. Safety and trust are present across nearly all research topics, shaping how questions are framed and investigated. This dominance extends beyond thematic distribution to methodological choices; quantitative studies based on models such as UTAUT dominate, while qualitative and mixed-methods approaches, better suited to exploring fairness, cultural dynamics, or governance processes, remain relatively scarce. Even within Theme F, explicitly labeled ‘Cross-Cultural and Urban Contexts,’ discussions often unfold indirectly through security, risk, or institutional arrangements, with cultural differences themselves rarely serving as the core analytical focus. Taken together, these findings respond directly to RQ1 and RQ2 by revealing the thematic concentration on safety and trust, while RQ3 highlights persistent empirical gaps in governance, equity, and cross-cultural embedding processes.

Regional distribution further reinforces these tendencies. Europe, North America, and Asia constitute the primary geographic sources of existing research. European studies tend to emphasize institutional design and policy environments, while Asian research has grown rapidly in volume but predominantly focuses on technological performance and system efficiency. In contrast, Latin America, the Middle East, and Africa are virtually absent from empirical research. This gap reflects both data and resource disparities and reveals the current research trajectory’s reliance on specific governance models and research infrastructure. Consequently, conclusions about ‘public acceptance’ are often based on experiences from a limited number of regions and specific urban types. At the technology type level, the literature shows a clear bias toward private AVs. Although SAVs and automated public transit are often viewed in policy and academic discussions as key pathways to achieving equity and sustainability goals, empirical research on them remains limited. This disconnect between ‘imagined potential’ and ‘empirical evidence’ constrains systematic comparisons of the social impacts of different automation pathways, leaving discussions of equity and sustainability largely confined to normative debates.

Overall, these patterns are not accidental, but are shaped by the interplay of research methodologies, practical contexts, and disciplinary structures. It is important to note that these observations reflect tendencies within the AV user acceptance literature rather than the entirety of autonomous mobility research. At the practical level, existing research heavily relies on technology pilots and early demonstration projects. Methodologically, quantitative models align more readily with safety and acceptance issues. Disciplinarily, the dominance of engineering and transportation economics has subtly reinforced a research orientation centered on efficiency and risk.

5.2. Theoretical Implications

This review’s comprehensive analysis indicates that AV research urgently requires an interdisciplinary perspective capable of connecting technology acceptance, social policy, and cultural context, rather than merely expanding existing models at the variable level. TAM, UTAUT and their derivative models demonstrate strong explanatory power in predicting individuals’ willingness to adopt autonomous vehicle technology. However, their theoretical focus remains confined to the individual level. Institutional conditions, social relationships, and spatial variations are often treated as exogenous contexts, while cultural factors are predominantly treated as control variables rather than as ongoing social processes that continuously shape technological meaning and usage practices. This partly explains why equity, community, and cultural issues struggle to become central analytical subjects in existing research. Conversely, studies grounded in macro-cultural dimensions (e.g., Hofstede) or technological transformation perspectives, while offering crucial conceptual frameworks, often operate at overly broad scales. This makes it difficult to explain the actual embedding processes of autonomous vehicle technology within specific cities, communities, and daily mobility practices. Although technological transformation theories emphasize systemic changes, they often overlook how the public forms attitudes and adapts its behavior under uncertainty. This points to AVs as a socio-technical phenomenon rather than a purely transportation technology issue. To fully understand AV acceptance and impacts, research must move beyond single-discipline approaches. A closer dialogue between transportation studies, social policy, urban research, and cultural analysis is needed. Research on AVs must shift from questioning ‘whether it will be accepted’ to examining ‘how it can be embedded within specific social and urban contexts,’ thereby redefining the relationship between technology adoption, social equity, and cultural diversity.

5.3. Practical/Policy Implications

From both practical and policy perspectives, this review suggests that relying solely on user acceptance or technological feasibility as the primary basis for AV deployment may overlook potential implications for equity and sustainability. Existing research indicates that AVs initially tend to serve groups with the means to pay and the capacity to adapt to new technologies. Without careful governance guidance, this could exacerbate existing transportation inequalities. Early consideration of equity objectives, through service coverage, pricing policies and alignment with existing public transit, may help guide implementation rather than treating these aspects as secondary concerns. This issue appears particularly relevant for medium-sized cities, which often have limited fiscal resources, relatively underdeveloped public transit provision, and residents who rely heavily on private vehicles.

Treating AVs as a mere replacement for existing transportation modes could risk overlooking institutional constraints and social conditions in practical operation. A more balanced approach is to consider AVs as a complementary public service, potentially filling temporal and spatial gaps in public transit coverage rather than accelerating privatized mobility patterns. Moreover, policies addressing older adults and other vulnerable groups remain limited. While studies often measure ‘acceptance’, this does not necessarily translate into improved access or mobility. If service design, interface complexity or trust-building mechanisms fail to account for the daily practices of these groups, automated transportation systems could inadvertently create new forms of exclusion. Consequently, policy formulation and system design would benefit from being grounded in specific usage scenarios and from incorporating the needs of vulnerable groups early in planning and throughout ongoing evaluation.

5.4. Future Research Directions

Based on the comprehensive analysis of this review, future research on AVs necessitates systematic adjustments in methodology, spatial scale, and research pathways. First, methodologically, it is essential to move beyond the predominant cross-sectional survey paradigm and increasingly adopt longitudinal mixed-methods designs at the community level. This approach better captures the temporal evolution of public attitudes, trust, and usage practices. Relying solely on one-time attitude measurements fails to reflect the social embeddedness of automated transportation in real urban contexts. Second, in terms of spatial scope, the existing literature relies heavily on a limited number of global cities and national-level studies, while medium-sized cities and non-core regions remain chronically neglected. These cities often exhibit both high car dependency and limited governance resources, resulting in automation transition pathways that differ significantly from those of large cities yet better reflect the realities of most urban areas. Regarding research approaches, there is a need to shift from technology and policy-centric top-down assessments toward more participatory, bottom-up research orientations. By incorporating the perspectives of residents and local practitioners, studies can gain deeper insights into how autonomous driving impacts daily travel decisions, community relations, and trust formation. Taking medium-sized European cities as an example, Galway is not an ‘exception’ but a representative research site where private car dependency coexists with sustainable transformation. Community-level studies in such cities, combining co-creation workshops, semi-structured interviews, travel diaries, scenario simulations, and long-term tracking, can systematically analyze the impact of automated transportation on accessibility for vulnerable groups, community-level sustainability outcomes, and the evolution of public trust. By reframing AVs not merely as a standalone technological object but as an urban system embedded within daily life and social relations, future research can provide a more robust empirical foundation for context-sensitive policy design. It will also deepen interdisciplinary research on autonomous vehicle technology, both in theory and methodology.

6. Conclusions

AVs are more than just a technology; they operate within communities and social structures. Looking at automated public transit over the past decade, research has focused mostly on safety and trust. Studies on community-level effects, fairness, or cultural factors are far less common. Public acceptance does not automatically mean access for everyone. Many studies examine individual attitudes or technology expectations. Few explore how AVs fit into daily travel or how they affect vulnerable groups, such as the elderly and low-income households. This gap is especially clear in medium-sized cities, where people rely heavily on private cars, but resources and institutional support are limited. Research needs to move beyond asking whether AVs will be accepted. Instead, it should examine how they can be embedded in communities. AVs are not simply a replacement for current transport. They have the potential to shape accessibility, social relations, and sustainable lifestyles. Therefore, this review paper calls for future research to adopt more community-level, longitudinal, and cross-cultural empirical approaches, focusing on the actual operational logic and social consequences of automated transportation across different urban types. By integrating technological advancement with local governance, community practices, and social equity concerns, automated public transit holds promise as a tool that supports sustainable and inclusive urban transformation rather than a technological variable that exacerbates existing inequalities. In this sense, the critical question for automated public transit is not technological maturity but whether cities and communities are prepared to coexist with it.

This review has several limitations. First, the literature selection was limited to peer-reviewed journal articles in English, with the core corpus retrieved using a structured Scopus search and additional studies incorporated through citation-based identification and iterative snowballing. While this approach enhances conceptual linkage within the AV user acceptance research network, it may underrepresent non-English scholarship, grey literature, and studies not connected through citation pathways. Second, no formal review protocol was preregistered; instead, the study relies on transparent retrospective documentation of eligibility criteria, coding definitions, keyword sets, and screening decisions. These records are provided in the Section 3 and Appendix A and Appendix B to allow full auditability of the review process. Predefined inclusion parameters were applied consistently throughout both the structured search and citation-based enrichment stages to maintain procedural coherence. Third, many included studies focus on pilot projects, early deployments, or simulations, which limits evidence on long-term or large-scale implementation, particularly for highly automated vehicles at Levels 4–5. Third, this review applied a structured thematic coding approach to map patterns in public acceptance, equity, governance, and social issues. The coding framework captures conceptual trends and research gaps rather than providing precise quantitative estimates. Future research could build on these findings through longitudinal, community-level, and cross-cultural empirical studies to examine how autonomous mobility systems are embedded in everyday practices and evolving urban contexts.