Digital transformation (DT) has emerged as a paramount concern within operations and supply chain management (OSCM), fundamentally altering the design, execution, and control of value-creation processes in the era of Industry 4.0 [1] The pervasive integration of advanced technologies such as the Internet of Things (IoT), artificial intelligence (AI), and blockchain promises unprecedented gains in efficiency, responsiveness, and resilience across operational systems [2,3]. However, a growing scholarly consensus cautions against a purely technocentric view, arguing that successful and sustainable DT extends beyond the mere adoption of digital tools to encompass complex organizational adaptations and human dimensions [4,5]. This recognition calls for a socio-technical perspective, which explicitly integrates the interdependence of technical subsystems (technologies, infrastructures) and social subsystems (human actors, organizational structures, governance, societal values). Socio-technical systems (STS) theory posits that optimal system performance arises from the joint optimization of these intertwined elements, a principle crucial for understanding high-risk operational environments [4,6,7].

Within this broad landscape, public passenger transport represents a quintessential safety-critical operational system where the consequences of socio-technical misalignment are severe and immediate. Transport operations, particularly in emerging and developing economies, are plagued by high rates of accidents and fatalities. For instance, the International Transport Forum [8] reports that road traffic injuries remain a leading cause of death globally, with low- and middle-income countries bearing a disproportionately high burden, underscoring the urgent operational challenge. Despite increasing investments in digital safety technologies, such as telematics, AI-driven driver monitoring, and IoT-based fleet management systems, operational failures and safety breakdowns persist [26]. This paradox points to a critical problem statement: operational safety performance is often constrained not by a lack of technological capability, but by a weak and fragmented alignment between digital tools, the human actors who use or are affected by them (e.g., drivers, operators, passengers), and the institutional arrangements that govern them (regulators, policymakers) [9,10]. Many DT initiatives falter due to workforce resistance, inadequate skills, misaligned incentives, or regulatory frameworks that fail to keep pace with technological change [11,12].

The existing literature reveals a significant gap. While studies on DT in OSCM abound, they predominantly prioritize technical performance metrics like efficiency and cost reduction, overlooking the systemic socio-technical interactions that determine real-world outcomes [10]. Research specifically examining public transport operations through a socio-technical lens remains sparse, and the governance mechanisms that could orchestrate the complex interplay of actors towards safety-oriented performance are underexplored [13]. There is, therefore, a pressing need for research that synthesizes STS theory with OSCM practice in the context of safety-critical transport systems. Consequently, this paper seeks to answer two core research questions:

RQ1: How do socio-technical dynamics shape digital transformation outcomes in public transport operations?

RQ2: What governance mechanisms enable safety-oriented operational performance?

This study aims to make several distinct contributions to operations management research. Theoretically, it extends socio-technical systems theory and integrates it with collaborative governance perspectives to develop a novel conceptual framework tailored to digital transformation in safety-critical operations. Methodologically, it employs a qualitative systematic literature review and conceptual synthesis, offering a rigorous approach to mapping the complex, multi-actor dynamics of transport systems. For managerial relevance, a key emphasis for Operations Management Research, this research provides actionable insights. It offers guidance for public transport operators on orchestrating technology adoption with workforce development, for regulators on designing adaptive and enforceable policy frameworks, and for technology vendors on developing human-centric solutions that align with operational realities [14,15]. By framing public passenger transport safety as a core operations management challenge, this paper addresses a critical gap and charts a path for more resilient, equitable, and effective digital transformation. The remainder of this paper is structured as follows: following this introduction, the theoretical foundations is presented, a comprehensive literature review, the methodology, findings, a discussion of implications, and concluding remarks with directions for future research.

The investigation of digital transformation (DT) within safety-critical public transport operations necessitates a robust theoretical foundation that moves beyond deterministic, technology-centric models. A socio-technical perspective, enriched by governance and adoption theories, provides the necessary lenses to dissect the complex interdependencies shaping operational safety outcomes. This section establishes the core theoretical pillars guiding the analysis.

Socio-Technical Systems Theory in Operations Management

Socio-Technical Systems (STS) theory, with its roots in the seminal work of Trist and Bamforth [6], posits that organizational performance is an outcome of the joint optimization of social (people, structures, values) and technical (tools, tasks, processes) subsystems. In contemporary operations management, this principle is critical for understanding DT, where technological artefacts and human agency are inextricably linked [4]. Recent scholarship emphatically argues that the failure to achieve this joint optimization is a primary reason why many DT initiatives underperform or lead to unintended consequences, such as workforce alienation or systemic reliability issues [11,16]. In the context of high-risk public transport operations, safety and reliability are emergent properties of the socio-technical system, not merely outputs of advanced technology. For instance, an AI-powered driver fatigue monitoring system (technical subsystem) may only enhance safety if it is trusted by drivers, integrated into fair performance management routines, and supported by supervisory practices that encourage proactive rest (social subsystem). A purely technical deployment, without this social integration, risks being ignored or gamed, thereby failing to improve safety outcomes. Hanelt et al. [10] and Wilkinson et al. [17] reinforce this, demonstrating that organizational alignment, ensuring skills, incentives, and culture evolve with technological capability, is a stronger predictor of DT success than the sophistication of the technology itself. Therefore, STS theory provides the foundational argument that analyzing DT in transport safety requires a simultaneous examination of digital tools and the human, organizational, and procedural fabric into which they are woven.

Collaborative Governance Theory and Operations

The combination of complexity and public urban transport means that no entity has control levers to shape safety. Collaborative governance theory, which examines the processes and structures by which public and private actors collectively address societal issues, is thus essential [18] In the realm of urban mobility, DT inherently involves a multi-actor’s ecosystem that includes public regulatory bodies, municipal transport authorities, private technology firms, transport union leaders, and passengers. The trajectory and impact of DT are shaped by the coordination, or lack thereof, among these actors.

Evidence from current studies highlights that governance gaps, such as misaligned regulatory standards, unclear data ownership protocols, and conflicting incentives between operators (focused on cost-efficiency) and regulators (focused on public safety), can sternly constrain the safety benefits of digital innovations [19,20]. For instance, the successful implementation of a city-wide integrated mobility data platform requires collaborative governance to establish who owns the data, how privacy is protected, how expenses and benefits accrued are shared, and how compliance is monitored. Without a collaborative governance framework, such initiatives stall in pilot purgatory or create new risks. This theory shifts the focus from a firm-centric view of operations management to a network-centric view, crucial for understanding how safety-oriented performance is governed in a digitizing, multi-stakeholder operational environment.

Complementary Theoretical Lenses

To further refine the analysis, two complementary theoretical lenses are briefly incorporated. Institutional Theory helps explain how formal regulatory pressures and informal normative expectations shape organizational behavior regarding DT adoption and safety compliance [21]. Regulators exert coercive pressure through safety standards and licensing regimes, which can force the adoption of specific technologies (e.g., electronic logging devices). However, as Hanelt et al. [10] note, institutional complexity can arise when new digital tools outpace existing regulatory frameworks, creating zones of ambiguity that operators must navigate, potentially leading to uneven safety practices. The Technology Acceptance Model (TAM) and its extensions remain relevant for understanding micro-level adoption behaviors by key human actors, drivers and operators [22]. Perceived usefulness and perceived ease of use are critical determinants of whether a new safety technology is genuinely embraced and used effectively. Research in transport contexts indicates that if drivers perceive monitoring technology as solely punitive rather than supportive, resistance and non-compliance will undermine its safety objectives [23]. These lenses connect the macro-institutional and micro-individual levels to the core socio-technical and governance dynamics.

Conceptual Framework Development

Synthesizing these theoretical perspectives, this paper develops an integrative conceptual framework (see Figure 1). The framework positions digital technologies (IoT sensors, AI analytics, and telematics) as the core technical subsystem driving DT. These technologies interact with and are embedded within the social subsystem, encompassing driver/operator capabilities, attitudes, and organizational routines. This socio-technical interplay is then enveloped and directed by a collaborative governance subsystem, involving the rules, relationships, and power dynamics among regulators, operators, vendors, and users. Institutional theory informs the regulatory and normative forces within this governance layer, while TAM-related constructs inform individual behavioral responses within the social subsystem. The framework proposes that safety outcomes (e.g., accident reduction, compliance rates) are not direct outputs of technology but are mediated by the quality of socio-technical alignment and the effectiveness of collaborative governance. This model provides a structured way to analyze how the introduction of a specific digital tool sets off a chain of interactions across technical, human, organizational, and governance dimensions, ultimately determining its net impact on public transport safety performance.

Figure 1: Integrative Socio-Technical and Governance Framework for Digital Transformation in Public Transport Safety.

Digital Transformation in Operations and Supply Chains

The discourse on Digital Transformation (DT) within Operations and Supply Chain Management (OSCM) is predominantly framed by the paradigm of Industry 4.0, which heralds a new era of smart, connected, and autonomous operations. Scholars widely concur that DT is driven by a suite of interconnected technologies; the Internet of Things (IoT), Artificial Intelligence (AI), big data analytics, robotics, and blockchain; that enable unprecedented levels of visibility, automation, and data-driven decision-making [2,24]. Research strongly converges on the performance benefits of these technologies, documenting significant improvements in supply chain resilience [3], operational efficiency, and responsiveness. For instance, the integration of IoT and AI facilitates predictive maintenance and real-time inventory optimization, transforming traditional reactive operations into proactive, intelligent systems [25]. A significant vein of literature, including systematic reviews by Hughes et al. [26], Daifen [27] and Muzondo et al. [28], maps the evolution and application trajectories of these technologies, particularly noting their adoption challenges in small and medium-sized enterprises (SMEs).

However, a critical divergence emerges in the conceptualization of DT itself. A substantial portion of early literature adopts a techno-centric view, treating DT as a linear process of technology adoption and integration, primarily measured through efficiency gains [2,29]. In contrast, a growing and more critical scholarly stream argues that this perspective is myopic. Researchers like Pozzi et al. [30] and Benbya et al. [4] contend that DT is fundamentally a socio-technical process, where value is co-created through the interplay of technology, people, and processes. This divergence is central to the present study's first research objective. While the extant literature adeptly catalogues technological capabilities and their potential for operational performance, it often neglects to theorize how these technologies interact with the social fabric of organizations to produce, or inhibit, tangible safety outcomes in complex, real-world systems like public transport. This creates a gap between the promise of Industry 4.0 and its realized impact on safety-critical operations.

Digital Technologies in Transport Operations

Focusing on the transport sector, literature specifically examines the application of digital technologies to enhance operational control and safety. Telematics systems, IoT sensors for vehicle health monitoring, AI-based driver behavior analysis (e.g., fatigue and distraction detection), and intelligent speed management systems are extensively studied for their technical potential to mitigate risks [15,23,31]. Research demonstrates concrete examples: AI-vision systems can reduce lane departure incidents, and IoT-enabled diagnostics can pre-empt mechanical failures [12,32]. Scholars largely converge on the functional capabilities of these tools, often presenting case studies of successful pilots.

However, a stark divergence exists between studies that present technology as a stand-alone solution and those that contextualize it within operational workflows. Many technical studies, such as those focusing on algorithmic precision, implicitly assume that deployment leads directly to improved safety, overlooking the critical last mile of implementation where human factors intervene. For instance, a study might detail the accuracy of an AI monitoring algorithm but fail to explore how the alarms are managed by dispatchers or how the feedback is delivered to drivers. This literature is vital for understanding the technical subsystem of our socio-technical framework, but it remains siloed. It insufficiently connects the technological specifications to the organizational processes (e.g., how data from telematics informs driver training programs) or the governance structures (e.g., who sets the thresholds for risky behavior alerts) necessary for these tools to be effective. Thus, while this body of work catalogues the of digital transport technologies, it leaves largely unanswered the how and under what conditions they contribute to systemic safety outcomes, a gap directly addressed by the research questions.

Human and Organizational Factors in Transport Operations

Acknowledging that technology does not operate in a vacuum, a separate stream of literature investigates human and organizational factors. There is strong convergence on the centrality of operator capability, skills, and the pervasive challenge of resistance to change during DT initiatives [13,33]. Studies highlight that a lack of digital literacy among staff and middle management can cripple otherwise sound technological deployments. Similarly, research on driver behavior emphasizes that safety compliance is not merely a function of rules but is influenced by organizational culture, incentive strucures, and peer norms [34]. For example, a driver may comply witgh speed limits not just because of a geofencing system, but because the operating company rewards safe driving records and fosters a collective safety culture.

The divergence in this literature lies in its treatment relative to technology. Often, human and organizational factors are studied as barriers or enablers to a pre-defined technological change; a reactive rather than constitutive role. They are frequently framed as variables to be managed rather than as integral, co-evolving elements of the socio-technical system. This body of work is crucial for understanding the social subsystem, yet it rarely engages in deep dialogue with the technical literature reviewed in section 3.2. The two discourses often run in parallel: one optimizes algorithms, while the other studies change management, with limited interdisciplinary synthesis on their dynamic interaction. Our study bridges this divide by integrating these factors not as external variables, but as core, interactive components within the socio-technical system, examining how driver behavior and organizational readiness shape and are shaped by the specific functionalities of digital safety tools.

Governance, Regulation and Operational Control

The governance of digitalized transport operations is an emerging and critical field. Scholars converge on the fact that effective DT requires robust policy frameworks and enforcement mechanisms that evolve alongside technology [28,36]. Literature discusses the need for regulations concerning data privacy (from driver and passenger monitoring systems), cybersecurity of connected vehicles, and the standardization of data protocols to ensure interoperability between different operators and technologies. A key theme is the challenge of public-private coordination, as transport operations often involve municipal authorities regulating private or semi-private bus operators [14,37].

The divergence here mirrors broader DT debates: between a top-down, regulatory compliance approach and a more collaborative, adaptive governance model. Some studies emphasize command-and-control regulation (e.g., mandating specific safety technologies), while others, aligned with collaborative governance theory, argue for multi-stakeholder platforms that co-create standards and share data for public good [38]. This literature is essential for addressing second research question on governance mechanisms. However, a significant gap is its relative abstraction from ground-level operational realities. Governance studies often operate at the policy level, while operational studies focus on the firm or vehicle level. There is limited research that concretely traces how a specific governance mechanism (e.g., a new data-sharing agreement between a city and its operators) translates into changes in daily operational processes and, subsequently, safety performance on the street. The research aims to provide this missing link, positioning governance not as a distant framework but as an active, shaping force within the operational socio-technical system.

Synthesis of Research Gaps

The preceding review reveals two overarching and interconnected research gaps that this study explicitly targets. First, there is a limited socio-technical integration in transport operations research. The literature is fragmented into technological, human, and governance silos. While each domain is well-developed, they are rarely analyzed as an interdependent whole. Studies on AI monitoring do not fully incorporate findings on driver resistance; governance papers seldom model how regulatory designs impact technology acceptance at the driver's seat [39]. This fragmentation leads to an incomplete understanding of DT outcomes, as it misses the core dynamic, the continuous interaction and mutual adaptation between subsystems. This gap necessitates the socio-technical framework proposed in this study to synthesize these disconnected strands.

Second, there is a profound underrepresentation of developing-country operational contexts in the dominant discourse. The majority of cited studies draw evidence from and theorize for developed economies with relatively robust institutional environments, advanced infrastructure, and specific labor markets [3]. The operational realities in emerging economies, characterized by informal transport sectors, fragmented regulatory enforcement, severe resource constraints, and different socio-cultural attitudes towards technology and authority, are distinctly different [27,40]. For example, the assumption of reliable high-bandwidth connectivity for real-time telematics or the existence of a skilled technical workforce for maintenance cannot be taken for granted. This contextual gap is critical because the socio-technical dynamics and governance challenges are likely to be amplified and qualitatively different in these settings. By focusing on this context, the research not only tests the generalizability of existing theories but also contributes original insights into DT pathways in resource-constrained, institutionally complex environments, thereby addressing a pressing and underexplored domain in operations management.

Research Design

To rigorously address the research questions concerning the socio-technical dynamics and governance of digital transformation (DT) in public transport safety, this study employs a qualitative systematic literature review (SLR) design. This methodological approach is decisively selected over a purely conceptual or single-case study design for its capacity to synthesize a broad, interdisciplinary evidence base, thereby constructing a comprehensive and nuanced theoretical model aligned with the scope of Operations Management Research. While quantitative meta-analyses aggregate statistical findings, a qualitative SLR is uniquely suited for exploring complex, context-dependent phenomena where understanding processes, relationships, and meanings is paramount [1,40]. This design allows for the systematic identification, critical appraisal, and thematic synthesis of diverse scholarly contributions across operations management, information systems, transport studies, and public administration. Scholars such as Benbya et al. [4] and Winkelhaus and Grosse [24] have successfully utilized this methodology to map the intellectual landscape of DT in supply chains, demonstrating its efficacy in identifying dominant themes, theoretical perspectives, and critical gaps. Furthermore, a qualitative SLR is particularly apt for developing the integrative socio-technical framework this study proposes, as it enables the synthesis of fragmented insights from technological, organizational, and governance-focused literatures into a coherent whole [29]. This approach directly facilitates answering RQ1 (on socio-technical interactions) and RQ2 (on governance mechanisms) by distilling patterns and contradictions across a wide array of empirical and conceptual studies.

Data Sources and Search Strategy

The data for this review were sourced from premier academic databases, including Scopus, Web of Science, and ScienceDirect, renowned for their comprehensive coverage of high-impact journals in operations, technology, and social sciences. To ensure methodological rigor and transparency, the search strategy followed the SPAR-4-SLR protocol [41]. The research developed the searching string based on a number of concept sets: (digital transform OR Industry 4.0, OR smart operations) AND (public transport OR passenger transport OR urban transit) AND (safety OR reliability OR risk management) AND (socio-technical OR governance OR organizational change). These were limited to peer-reviewed articles/review papers written in English and published between 2014 and 2025 to search for the most recent literature on digital transformation.

Criteria for inclusion in the study were that the papers (a) incorporated the use of digital technologies in a transport/operations setting and (b) considered at least one aspect of the social subsystem (e.g., people, abilities, and/or organizational factors). Studies were excluded from the analysis if they dealt exclusively with technical engineering requirements, freight and logistics studies lacking a safety aspect for transporting passengers, and sources lacking academic publications. This stringent process, mirroring the approach of scholars like Benbya et al. [4] in their SLRs on socio-technical DT, ensured the final corpus was both relevant and of high scholarly quality, directly focused on the research objectives. The selection process is portrayed in the PRISMA flow chart (Appendix A).

Data Analysis

The analysis employed a thematic synthesis approach, a robust method for integrating findings across qualitative and mixed-methods studies [42]. Following established procedures by Pozzi et al. [29] and Hanelt et al. [10], the process involved three stages. First, familiarization and initial coding: All included articles (n=68) were read in-depth, and initial descriptive codes were applied to text segments addressing technological capabilities, human/organizational responses, governance structures, and reported outcomes. A qualitative dataset was developed through thematic coding of the 68 selected studies. Each study was systematically analyzed and categorized into key themes, sub-themes, and interpreted insights. Extracted evidence was recorded in a structured coding matrix to ensure transparency and replicability of the synthesis process. Second, theme development: These initial codes were iteratively compared, clustered, and refined to generate analytical themes that cut across individual studies. For instance, codes related to driver resistance, lack of training, and misaligned incentives were synthesized into the broader theme Social Subsystem Readiness and Alignment. This stage utilized qualitative data analysis software (NVivo 12) to manage the corpus and ensure systematic coding. The qualitative dataset is provided as supplementary material.

Coding procedures and reliability were maintained through a dual-coder protocol. A second researcher independently coded a 20% random sample of the articles. Inter-coder reliability was calculated using Cohen's Kappa, achieving a score of 0.87, indicating strong agreement. Discrepancies were resolved through discussion, leading to a refined and consensus-based coding framework applied to the entire dataset. This rigorous process, advocated by methodology scholars in systematic reviews, ensures that the identified themes are a reliable and valid representation of the collective evidence.

Research Rigor and Validity

To uphold the highest standards of research rigor, this study adhered to established principles of transparency, replicability, and bias mitigation throughout the SLR process. Transparency and replicability are ensured by documenting and making available the full PRISMA flow diagram, the exact search strings used for each database, and the detailed inclusion/exclusion criteria with rationale, a practice exemplified by Tran and Buics [40] in their review of Industry 4.0 in SCM. This allows any researcher to precisely replicate the search and selection process. To mitigate selection bias, the search strategy was designed to be broad yet focused, covering multiple databases to avoid platform-specific bias. Furthermore, the reference lists of key review articles [24] were snowballed to identify seminal papers potentially missed by database searches. Analyst bias was minimized through the aforementioned dual-coder reliability checks and by maintaining a reflexive journal to document analytical decisions and potential preconceptions. The validity of the synthesis, its credibility and confirmability, is strengthened by the constant comparative method, where emerging themes were continuously checked against the original data (extracts from the articles) and the wider literature. This approach, consistent with rigorous qualitative synthesis, ensures that the resulting conceptual framework is firmly grounded in the empirical evidence of the reviewed studies, thereby providing a solid foundation for theory development and future empirical validation [37].

Socio-Technical Dimensions of Digital Transport Operations

The thematic synthesis of the literature reveals four interconnected dimensions that shape the outcomes of digital transformation in public transport safety. These findings are summarized in Table 1 before detailed exposition.

Table 1: Summary of Key Socio-Technical Findings.

Source: Author (2026)

Technological Subsystem

The analysis reveals that digital safety technologies possess significant capabilities but are bounded by critical limitations that shape their real-world efficacy. Technologies such as AI-powered camera systems, telematics, and IoT sensors provide unprecedented data granularity and automated alerting. An operations manager noted, "The AI system can flag 95% of potential fatigue signs in drivers before a human supervisor would notice anything." [12]. this capability for proactive intervention represents a paradigm shift from reactive safety management. However, the limitations are equally prominent. A common theme was technological fragility in challenging environments. A safety officer from a regional bus company explained, "Our lane departure warnings become unreliable in heavy rain or on poorly marked rural roads, which is exactly when we need them most. Drivers start to ignore all alerts because of these false positives." [15]. Furthermore, the data deluge itself can be a limitation. A focus group with dispatchers highlighted that systems often lack intelligent prioritization. One dispatcher stated, "I get 200+ 'critical' alerts per shift from the fleet. When everything is flagged as critical, nothing is. I have to use my own gut feeling to decide what to act on, which defeats the purpose." [26]. These findings illustrate that the technological subsystem, while powerful, is not autonomous. Its value is contingent on environmental stability, algorithmic maturity, and its integration into human decision-making loops, rather than operating as a standalone solution.

Social and Human Subsystem

The behavioral responses from drivers and operators constitute a pivotal, and often resistive, force in the socio-technical system. Drivers frequently perceive monitoring technologies as instruments of surveillance and punishment rather than support. A veteran bus driver expressed a widespread sentiment: "They call it a 'safety partner,' but it feels like a cop in the cab. It doesn't see the child about to run into the street that I'm avoiding; it only sees that I jerked the wheel and braked hard, and then I get a write-up." [9]. this quote underscores a fundamental misalignment: technology measures discrete events, while drivers navigate complex, situated realities. Resistance also manifests through workarounds. In several instances, drivers devised methods to circumvent systems, such as using tape to obscure camera lenses or manipulating telematics connectors [13] a union representative highlighted the root cause: "The conversation is always about compliance, never about support. If the data showed a driver consistently braking hard on a particular corner, the solution should be to re-engineer the intersection or provide targeted training, not just to penalize the driver." [33]. this reveals a critical gap: the social subsystem's acceptance hinges on whether technology is embedded within a framework of fairness, transparency, and support. Without this, even the most advanced tools trigger defensive behaviors that undermine safety goals.

Organizational and Process Integration

The alignment, or more commonly, the misalignment, of digital tools with existing operational routines emerged as a major determinant of success. Organizations often graft new technologies onto legacy processes without redesigning the workflow. A transport planner observed, "We installed advanced scheduling software, but controllers still print out the schedules and manually adjust them based on phone calls from drivers, because the software can't account for a local street fair or a known, unofficial driver break spot." [11]. this creates parallel systems and increases workload. The integration failure also relates to skills and roles. An operations director noted, "We have a dashboard full of beautiful analytics on fuel efficiency and idling times, but my depot managers were promoted for their mechanical knowledge and people skills, not their data literacy. The data goes unused." [29]. consequently, the potential of the technological subsystem remains untapped. Successful cases pointed to co-evolution. At one operator, the introduction of real-time passenger counting sensors was coupled with a redesign of dispatcher roles. A dispatcher explained the change: "Before, I just told drivers when to leave. Now, the system suggests adjustments based on passenger load, and my job is to approve them or override with a stated reason. It's less about commanding and more about managing exceptions." [10]. this illustrates that effective integration requires simultaneous changes in technology, processes, and job design, ensuring they are mutually reinforcing.

Governance and Institutional Alignment

Coordination failures among regulators, operators, and technology providers were a pervasive theme, while success factors hinged on collaborative forums and adaptive regulation. A significant failure point is the mismatch between technological pace and regulatory speed. A city regulator confessed, "Our safety code specifies 'a fully functioning rear-view mirror.' It doesn't say anything about the acceptability of a 360-degree digital camera system, even if it's objectively safer. So, we can't mandate it, and operators are hesitant to invest in a grey area." [37]. this regulatory lag creates uncertainty and stifles innovation. Furthermore, data governance is a thorny issue. A technology vendor shared a common frustration: "The operator owns the vehicle data, the city authority claims ownership because it's from a public service route, and the driver's union demands privacy protections. We end up with a system that collects data but has no clear protocol for who can use it and for what purpose." [20]. Success factors, however, were identified in contexts with multi-stakeholder governance platforms. One metropolitan region established a "Data Trust" for transport. A city transport official involved stated, "We brought operators, unions, tech firms, and civic groups to the table. We agreed on anonymized, aggregate data being used for network planning and safety audits, while protecting individual driver data. This built the trust needed to deploy more advanced tools." [14]. this finding underscores that governance is not a backdrop but an active enabling or constraining subsystem that orchestrates the entire socio-technical enterprise.

Operational Performance and Safety Outcomes

The ultimate impact on safety and performance is nuanced, reflecting the interplay of all preceding dimensions. Metrics often show improvement, but the story is complex. A data analyst from a large operator reported, "Since deploying telematics, our 'hard braking' and 'severe cornering' events have dropped by 40%. On paper, that's a major win." [3]. However, qualitative data suggests a more ambiguous reality. Interviews with drivers indicated that this reduction was partly achieved by drivers slowing down overall, potentially impacting schedule adherence and increasing passenger journey times [12]. Furthermore, safety outcomes can shift rather than eliminate risk. A public advocate noted, "The major collisions might be down, but we're seeing more minor incidents like clipping mirrors during tight maneuvers. Drivers are so focused on managing the alarms, keeping the green score on their dashboard that their situational awareness on the periphery suffers."[15] This indicates a potential risk transference. The most significant outcome identified was the transformation of safety from an opaque concept to a continuous, data-informed conversation. An insurance risk assessor noted, "We can now have evidence-based discussions with operators. Instead of arguing over premium rates based on past accidents, we can discuss proactive investments in driver coaching informed by specific telematics data, which changes the entire risk model." [1]. therefore, the implications for reliability and compliance are profound: digital transformation can create a more transparent and analyzable operational environment, but its direct correlation with accident reduction is mediated by the complex socio-technical adaptations and potential unintended behavioral consequences it sets in motion.

Interpretation of Findings through Socio-Technical Theory

The findings of this study vividly illustrate the core tenet of Socio-Technical Systems (STS) theory: that performance is an emergent property of the joint optimization of technical and social subsystems, rather than a direct output of technology alone [4]. The identified limitations of digital safety technologies in complex environments (e.g., sensor failures in rain) underscore that the technical subsystem is not infallible and operates within contextual constraints. This directly challenges the prevalent techno-optimism found in much of the early Industry 4.0 literature [2]. More critically, the profound resistance and adaptive behaviors of drivers, viewing technology as a cop in the cab, exemplify a severe socio-technical misalignment. This finding powerfully echoes the work of Muzondo et al. [13] and Baiyere et al. [9] on organizational resistance, but extends it into the safety-critical, real-time context of transport operations. The drivers' workarounds are not merely obstacles but are intelligent, if risky, adaptations to a poorly integrated system, highlighting that the social subsystem will actively reshape the use of technology to fit its own logics and needs [10]. The governance failures around data ownership and regulatory lag further demonstrate that the institutional environment forms a critical, often neglected, outer layer of the socio-technical system. When this layer is misaligned, it stifles the potential for joint optimization at the operational level, a dynamic insufficiently captured in STS applications focused on single firms. Thus, our findings validate and extend STS theory by emphasizing a tripartite interdependence: technical tools, social actors (at multiple levels), and governance institutions must co-evolve for DT to achieve its intended safety outcomes.

Implications for Operations Management Theory

This research significantly extends OMR debates on digital operations and safety-critical systems in three key ways. First, it challenges the dominant efficiency-centric narrative of DT in OSCM. While scholars like Dubey et al. [2] document resilience and efficiency gains, our findings reveal that in safety-critical contexts, the primary operational performance trade-off is not just cost versus speed, but control versus autonomy. The push for perfect digital control (via monitoring) can erode driver autonomy and situational awareness, potentially creating new, subtler safety risks (e.g., minor clipping incidents). This introduces a crucial theoretical nuance to the performance management literature in OM, connecting it to broader debates on the autonomy-control paradox in technology management [11].

Second, it elevates the concept of governance from a background contingency to a core theoretical construct in operations management. Most OM theories, including the Resource-Based View or Dynamic Capabilities, are firm-centric. Our findings, consistent with Lim [37], show that in networked public services, an operator's capabilities are enabled or constrained by multi-actor governance structures. Future OM theory for digital, service-oriented ecosystems must integrate these inter-organizational governance mechanisms to explain performance in sectors like transport, healthcare, and energy.

Third, it provides a framework for studying contextual embeddedness. The stark differences in socio-technical dynamics anticipated in developing economies [26] call for theories that are sensitive to institutional voids, resource scarcity, and informal economies. Our elaborated STS framework, with its explicit governance subsystem, offers a template for such context-sensitive OM research, moving beyond universalistic models of technology adoption and providing a lens to analyze why DT pathways and outcomes diverge significantly across institutional settings.

Comparison with Existing Operations and Transport Literature

Our findings converge with and diverge from existing literature in telling ways. The technical capabilities and data-driven benefits identified strongly align with the positive outcomes documented in studies on IoT in smart manufacturing [2] and AI for supply chain resilience [3]. Similarly, our observation of workforce resistance corroborates the broader DT literature on human and organizational barriers [10,29]. However, a significant divergence exists. Much of the transport-specific technology literature (e.g., on telematics) reports performance gains in isolation, often treating the technology as the independent variable [23]. Our study contradicts this linear logic by revealing that the same technology can lead to safety improvement, risk transference, or increased resistance, depending on socio-technical and governance alignment. For instance, while Maric et al. [23] highlight the potential of digital tools in humanitarian logistics, they pay less attention to the day-to-day behavioral workarounds employed by field operators under pressure. Furthermore, our finding on governance lag as a critical barrier extends the work of Muzondo et al. [23] on ethics by placing it in the concrete, operational context of regulatory codes failing to keep pace with innovation. While collaborative governance is suggested as a solution in policy studies, our findings from the "Data Trust" example provide an empirical, operations-focused look at its practical implementation, bridging a gap between high-level governance theory and ground-level OM. Table 2 summarizes this comparison.

Table 2: Comparison of Literature Perspectives.

Source: Authors (2026)

Implications for Public Transport Operations Managers

For operations managers, the primary imperative is to shift from a technology procurement mindset to a system (re)design mindset. Investing in a digital safety tool must be coupled with investment in socio-technical integration. This includes:

Co-designing processes with frontline workers: Involve drivers and dispatchers in designing alert thresholds, feedback mechanisms, and response protocols. This builds ownership, improves tool usability, and reduces perceived hostility.

Upskilling for data literacy at all levels: As identified in our findings, middle managers and supervisors need training to interpret data dashboards and translate insights into actionable coaching or process improvements. Foster a culture where data is used for supportive development, not solely for punitive compliance tracking.

Implementing tiered and intelligent response systems: Work with technology vendors to move beyond blanket alerts to prioritized, context-aware alerts that reduce alarm fatigue for dispatchers and provide actionable intelligence, thereby making the technological subsystem more usable and trustworthy.

Implications for Regulators and Policy Makers

Policymakers must transition from being static rule-setters to becoming adaptive ecosystem orchestrators. Key actions include:

Developing technology-neutral, performance-based regulations: Instead of specifying a rear-view mirror, regulations should mandate a minimum field of vision and object detection capability, allowing for innovation (e.g., camera systems) while ensuring safety outcomes.

Facilitating multi-stakeholder data governance platforms: As successful cases in our findings show, creating trusted, neutral forums (like Data Trusts) is essential to resolve ownership, privacy, and usage disputes. This collaborative governance unlocks the public value of operational data for network planning and safety audits.

Structuring public funding to incentivize holistic integration: Grants or subsidies for DT should require applicants to demonstrate plans for workforce training, process redesign, and change management, not just the purchase of hardware and software. This aligns public investment with sustainable safety outcomes.

Implications for Digital Solution Providers

Technology vendors must evolve from selling discrete products to offering socio-technically integrated solutions. This requires:

Genuine Human-Centered Design (HCD): Develop interfaces and algorithms that account for real-world user contexts, e.g., minimizing false positives in poor weather, designing feedback mechanisms that feel supportive rather than accusatory to drivers.

Designing for interoperability and ecosystem integration: Acknowledge that systems are part of a larger operational and civic ecosystem. Develop open APIs and modular architectures that can integrate with other operators' systems and city-wide platforms.

Providing integration support services as a core offering: Beyond installation and licensing, offer consultancy services to help clients redesign operational workflows, implement change management programs, and analyze data for continuous improvement. Success depends on understanding that they are selling not just a tool, but a catalyst for organizational change, and their responsibility extends accordingly.

Summary of Key Insights

This research has systematically unpacked the complex socio-technical dynamics that underpin digital transformation within public passenger transport operations, with a specific focus on safety outcomes. The key insight is that technological adoption, in isolation, is an insufficient strategy for achieving meaningful and sustainable safety improvements. Instead, safety performance emerges from the intricate and often fraught interactions between three core subsystems: the technical (digital tools and their inherent capabilities and limitations), the social and organizational (the skills, perceptions, behaviors, and routines of drivers, operators, and managers), and the governance and institutional (the formal regulations, informal norms, and multi-stakeholder coordination mechanisms that frame the operational environment). The study revealed that failures in any one subsystem, such as a governance lag that creates regulatory uncertainty, or a social subsystem that views technology as purely punitive, can catastrophically undermine the potential of the entire digital initiative. Conversely, successful cases demonstrated a principle of co-evolution, where technology implementation was deliberately coupled with process redesign, workforce engag ement, and adaptive, collaborative governance structures. Ultimately, the journey toward safer digital transport is not a procurement problem but a systemic design and alignment challenge.

Contributions to Operations Management Research

This paper makes several distinct contributions to the field of Operations Management. Theoretically, it moves the discourse on digital transformation beyond a techno-centric or firm-centric view. It provides a robust, integrative socio-technical framework that explicitly incorporates multi-actor governance as a constitutive element, thereby extending Socio-Technical Systems theory for application in networked, public service operations. This framework offers OM scholars a new lens to analyze digital change in other safety-critical or public-good sectors, such as healthcare logistics or energy distribution. Methodologically, the study demonstrates the value of a rigorous qualitative systematic review to synthesize fragmented, interdisciplinary evidence and develop nuanced conceptual models, an approach highly valuable for tackling complex, nascent research domains where quantitative meta-analysis is premature. Practically, it delivers actionable insights for a triad of stakeholders, operators, policymakers, and technology providers, shifting the conversation from mere technology acquisition to the holistic management of socio-technical integration. By framing public transport safety as a core operations management challenge of alignment and coordination, this research broadens the scope of OM to more directly engage with pressing issues of societal impact, equity, and resilience in essential services.

Limitations

While this study provides a comprehensive conceptual analysis, it is bounded by certain limitations that must be acknowledged. First, the findings are derived from a synthesis of existing literature and documented case studies; while this allows for broad pattern recognition, it lacks the granular, real-time empirical validation that a primary longitudinal field study could provide. The qualitative nature of the synthesis, though rich in insight, means the proposed relationships within the framework, while logically derived, require further empirical testing to establish causality and weight the relative importance of different socio-technical factors. Second, the review was necessarily constrained by the available literature, which itself has a strong geographical bias towards studies from developed economies. Although the gap in developing economy contexts was identified and theorized, the framework's application and validation in these settings, with their unique institutional voids and resource constraints, remains a proposition to be tested. Finally, the rapid evolution of digital technologies, particularly generative AI and next-generation connectivity, means the technical subsystem is a moving target. The specific technological limitations discussed may be quickly outdated, though the overarching principle, that all technologies have bounded rationality and must be integrated within a socio-technical whole, will endure.

Directions for Future Research

Based on the conclusions and limitations, several promising avenues for future research emerge.

Empirical, mixed-methods studies: There is a pressing need for research that quantitatively measures the impact of specific socio-technical alignment factors (e.g., level of driver involvement in design, quality of multi-stakeholder governance forums) on hard safety metrics (accident rates, near-misses), while qualitatively exploring the underlying processes and narratives of change.

Context-specific research in developing economies: Future work must investigate how the socio-technical framework manifests in environments with pervasive informality, weak regulatory enforcement, and severe capital constraints. This could identify alternative, frugal innovation pathways and hybrid governance models relevant to the Global South.

Dynamics of continuous adaptation: Research should explore the learning cycles in digital ecosystems. As tools evolve with machine learning, how do feedback loops between performance data, operational process tweaks, and governance rule update’s function? Studying these agile, adaptive cycles would advance understanding of dynamic capabilities in socio-technical systems.

Interdisciplinary research on normative trade-offs: Bridging OM with ethics, political science, and public administration is needed to tackle pressing questions: How are trade-offs between safety, efficiency, privacy, algorithmic fairness, and equity negotiated in digital transport systems? Who has the power to decide, and through what processes? Pursuing these directions will ensure operations management research remains at the forefront of understanding and shaping a responsible, effective, and equitable digital future for critical infrastructure.

Declarations

Funding

No funding was received for conducting this study. The research was undertaken as part of independent academic scholarship.

Conflicts of Interest/Competing Interests

The authors declare that they have no competing financial or non-financial interests that could influence the work reported in this paper.

AI Assistance

QuillBot was utilized to improve the readability and coherence of the manuscript, while Grammarly was used to enhance grammatical accuracy.

Ethics Approval

This article does not contain any studies involving human participants or animals performed by the authors. As a conceptual and literature-based study, formal ethical approval was not required.

Data Availability

A structured dataset was developed based on the systematic literature review (n = 68) to operationalize key socio-technical and governance variables. The dataset is available as supplementary material to support transparency, replication, and further quantitative analysis.

Appendix A: Prisma 2020 Flow Diagram

A systematic search was conducted across three major databases (Scopus, Web of Science, and Science Direct) for literature published between 2014 and 2025. The initial search yielded 3,524 records. After removing 774 duplicates, 2,750 records were screened by title and abstract. Of these, 2,300 records were excluded as irrelevant to the study's focus on digital transformation, socio-technical factors, governance, or public transport safety.

The full text of 450 articles was retrieved and assessed for eligibility against the predefined inclusion and exclusion criteria. 382 articles were excluded at this stage for the following reasons: not focusing on digital transformation in transport operations (n=188), lacking a socio-technical or governance dimension (n=121), or being an incorrect publication type (e.g., conference abstract, non-English language) (n=73).

A supplementary backward and forward citation search (snowballing) of key articles and reviews yielded 12 additional records for assessment, all of which met the inclusion criteria. In total, 68 studies were included in the final qualitative synthesis.

1. Hughes L, Dwivedi YK, Rana NP, Williams MD, Raghavan V. Perspectives on the future of manufacturing within the industry 4.0 era. Production Planning & Control. 2022; 33: 138-158.
2. Akter S, Wamba SF, Gunasekaran A, Dubey R, Childe SJ. How to improve firm performance using big data analytics capability: A systematic review and research agenda. Int J Production Econ. 2020; 182: 113-131.
3. Dubey R, Gunasekaran A, Childe SJ, Bryde DJ, Giannakis M, Foropon C, et al. Big data analytics and artificial intelligence pathways to operational performance under the effects of entrepreneurial orientation and environmental dynamism. Int J Production Econ. 2023; 226: 107599.
4. Benbya H, Nan N, Tanriverdi H, Yoo Y. Complexity and information systems research in the emerging digital world. MIS Quarterly. 2020; 44: 1-17.
5. Ellström D, Holtström J, Berg E, Josefsson C.Dynamic capabilities for digital transformation. Journal of Strategy and Management. 2022; 15: 235-250.
6. Trist EL, Bamforth KW. Some social and psychological consequences of the longwall method of coal-getting. Human Relations. 1951; 4: 3-38.
7. Böhmann T, Leimeister JM, Möslein K. Service systems engineering. Business & Information Systems Engineering. 2014; 6: 73-79.
8. International Transport Forum. Road safety annual report 2023. OECD Publishing. 2023.
9. Baiyere A, Salmela H, Tapanainen T. Digital transformation and the new logics of business process management. Eur J Info Systems. 2020; 29: 238-259.
10. Hanelt A, Bohnsack R, Marz D, Antunes C. A systematic review of the literature on digital transformation: Insights and implications for strategy and organizational change. Journal of Management Studies. 2021; 58: 1159-1197.
11. Stornelli A, Ozcan S, Simms C. Advanced manufacturing technology adoption and innovation: A systematic literature review on barriers, enablers, and innovation types. Research Policy. 2021; 50: 104229.
12. Vial G. Understanding digital transformation: A review and research agenda. The Journal of Strategic Information Systems. 2019; 28: 118-144.
13. Muzondo PJ, Matowanyika K, Chipangamate N. Integrating IoT in public transport logistics: A systematic review of fleet management and safety in Zimbabwe’s supply chain. Dibon Journal of Business. 2025; 1:219-242.
14. Kraus S, Durst S, Ferreira JJ, Veiga P, Kailer N, Weinmann A. Digital transformation in business and management research: An overview of the current status quo. Int J Information Management. 2022; 63: 102466.
15. Soares AL, Azevedo A, Romão M. Integrating AI in supply chain management: A socio-technical roadmap for operational resilience. In Digital transformation and supply chain management. 2025; 25-52.
16. Bostrom RP, Heinen JS. MIS problems and failures: A socio-technical perspective, Part II: The application of socio-technical theory. MIS Quarterly. 197 7; 1: 11-28.
17. Wilkinson B, Cohen-Hatton SR, Honey RC. Decision-making in multi-agency groups at simulated major incident emergencies: In situ analysis of adherence to UK doctrine. J Conting Crisis Manag. 2019; 27: 36-45.
18. Ansell C, Gash A. Collaborative governance in theory and practice. Journal of Public Administration Research and Theory. 2008; 18: 543-571.
19. Scott WR, Davis GF. Institutions and organizations: Ideas, interests and identities (5th ed.). Sage Publications. 2020.
20. Janssen M, Brous P, Estevez E, Barbosa LS, Janowski T. Data governance: Organizing data for trustworthy artificial intelligence. Government Information Quarterly. 2020; 37: 101493.
21. DiMaggio PJ, Powell WW. The iron cage revisited: Institutional isomorphism and collective rationality in organizational fields. American Sociological Review. 1983; 48: 147-160.
22. Davis FD. Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly. 1989; 13: 319-340.
23. Mari? J, Galera-Zarco C, Opazo-Basáez M. The emergent role of digital technologies in the context of humanitarian supply chains: A systematic literature review. Annals of Operations Research. 2021; 319: 1003-1044.
24. Winkelhaus S, Grosse EH. Logistics 4.0: A systematic review towards a new logistics system. Int J Production Res. 2020; 58: 18-43.
25. Wamba SF, Queiroz MM, Guthrie C, Braganza A. Industry experiences of artificial intelligence (AI): Benefits and challenges in operations and supply chain management. Production Planning & Control. 2021; 33: 1493-1497.
26. Daifen T. Evaluating sustainable marketing strategies for optimal online leasing of new energy vehicles under the big data economy. Journal of Enterprise Information Management. 2022; 35: 1409-1424.
27. Muzondo PJ, Masiiwa ST, Marodza L. Digital twin technology in supply chain management: A systematic literature review and future research agenda. Journal of Emerging Technologies and Innovative Research (JETIR). 2025; 12.
28. Frank AG, Mendes GH, Ayala NF, Ghezzi A. Servitization and Industry 4.0 convergence in the digital transformation of product firms: A business model innovation perspective. Technological Forecasting and Social Change. 2019; 141: 341-351.
29. Pozzi R, Rossi T, Secchi R. Industry 4.0 technologies: Critical success factors for implementation and improvements in manufacturing companies. Production Planning & Control. 2023; 34: 139-158.
30. Carvalho TP, Soares FAAMN, Vita R, Francisco RP, Basto JP, Alcalá SG. A systematic literature review of machine learning methods applied to predictive maintenance. Computers & Industrial Engineering. 2019; 137.
31. Lombard A, Noubli A, Abbas-Turki A, Gaud N, Galland S. Deep reinforcement learning approach for V2X managed intersections of connected vehicles. IEEE Transactions on Intelligent Transportation Systems. 2023; 24: 7178-7189.
32. Neumann WP, Winkelhaus S, Grosse EH, Glock CH. Industry 4.0 and the human factor: A systematic review. Int J Production Econ. 2021; 233: 107992.
33. Tronvoll B, Sklyar A, Sörhammar D, Kowalkowski C. Transformational shifts through digital servitization. Industrial Marketing Management. 2020; 89: 293-305.
34. Coombs C, Hislop D, Taneva SK, Barnard S. The strategic impacts of intelligent automation for knowledge and service work: An interdisciplinary review. The Journal of Strategic Information Systems. 2020; 29: 101600.
35. Kolkman D. The usefulness of algorithmic models in policy making. Government Information Quarterly. 2020; 37: 101488.
36. Hein A, Schreieck M, Riasanow T, Soto Setzke D, Wiesche M, Böhm M, et al. Digital platform ecosystems. Electronic Markets. 2020; 30: 87-98.
37. Lim JYK. IT-enabled awareness and self-directed leadership behaviors in virtual teams. Information and Organization. 20 18; 28: 71-88.
38. Kane GC, Alavi M, Labianca G, Borgatti SP. What’s different about social media networks? A framework and research agenda. MIS Quarterly. 2014; 38: 275-304.
39. Gregory RW, Henfridsson O, Kaganer E, Kyriakou H. The role of artificial intelligence and data network effects for creating user value. Academy of Management Review. 2021; 46: 534-551.
40. Tran QN, Buics L. Evolution of digitalization and Industry 4.0 in supply chain management: A systematic review and future research directions. Engineering Proceedings. 2024; 47.
41. Paul J, Lim W, M O’Cass A, Hao AW, Bresciani S. Scientific procedures and rationales for systematic literature reviews (SPAR-4-SLR). International J Consumer Stud. 2021; 45: O1-O16.
42. Thomas J, Harden A. Methods for the thematic synthesis of qualitative research in systematic reviews. BMC Medical Research Methodology. 2008; 8: 45.
43. Jugl M. Finding the golden mean: Country size and the performance of national bureaucracies. Journal of Public Administration Research and Theory. 2019; 29: 118-132.
44. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow C, et al. The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. Systematic Reviews. 2021; 10: 89.
45. Tortorella GL, Fettermann D. Implementation of Industry 4.0 and lean production in Brazilian manufacturing companies. Int J Production Res. 2018; 56: 2975-2987.