With one foot in the future – Situating autonomous systems in mining through communities of practice

In recent years, autonomous systems have become an integral part of the modern mining industry. These systems take various forms, but one of the most widespread and tangible examples is the use of autonomous vehicles for transporting ore and waste rock (Long et al., 2024; Rogers et al., 2019). While such vehicles have been under development and in use for some time (Walker, 2014) – they are not a novelty in and of themselves – their adoption is now gaining broader traction across the industry (Leonida, 2022). No longer limited to experimental settings, autonomous vehicles are increasingly integrated into everyday operations. Over time, these systems have been refined, becoming more sophisticated and capable of handling complex and dynamic functioning (Morton, 2020). This ongoing evolution marks a shift from earlier forms of automation, which targeted discrete functions, to autonomous systems with broader reach and growing capacity to act independently – interpreting their environment, making decisions, and adapting accordingly. With the introduction of emerging artificial intelligence tools, we expect that autonomous systems in mining will become increasingly self-governing – capable of handling more complex tasks with even less human intervention. As these systems gain the capacity to make decisions and act independently, they not only reshape work processes, but may also begin to subtly shift the balance of agency in human-technology relations, prompting new questions about how control, responsibility, and trust are negotiated in increasingly automated work environments.

Yet, this transformation does not take place in a vacuum. It unfolds within complex social systems – in research projects, at conferences among representatives of different companies and interest organisations, within internal development initiatives at mining companies, and not least on the shop floor. In these contexts, the introduction of autonomous technology is not merely a technical deployment, but a process of negotiation, interpretation, meaning-making, and situated learning. It is enacted by meaning-making beings – not only rational agents but socially situated practitioners – who make sense of and adapt to new technologies in relation to their existing practices and shared understandings. The implementation is carried out by trained professionals – experts in mining operations as well as engineers and technical specialists – who bring together domain knowledge and technological expertise to translate autonomous systems into workable solutions on the ground. Similar insights have been highlighted in studies of organisational change, which show that technology becomes embedded not simply through technical adjustments but through collective sensemaking and the negotiation of new routines (Tootoonchy and Clegg, 2025), and that successful adoption often relies on stakeholder engagement, continuous training, and distributed forms of leadership to align technological change with organisational practices (Oguntoyinbo et al., 2025). In this view, the capacity of an organisation to adapt rests as much on social processes of interpretation and coordination as on technical performance.

A range of theoretical traditions offer important lenses for understanding how new technologies become embedded in organisational life. Sociotechnical systems theory emphasises the joint optimisation of social and technical subsystems (Baxter and Sommerville, 2011; Mumford, 2006), while human–machine interaction research highlights the cognitive, ergonomic, and safety-related dimensions of operating automated systems (Parasuraman and Riley, 1997; Sheridan and Parasuraman, 2005). Likewise, technology-adoption models such as TAM and UTAUT focus on individual attitudes, expectations and intentions as determinants of acceptance (Fred, 1987; Venkatesh et al., 2016). Taken together, these theories illuminate important aspects of technology change but leave less visible the ongoing, collective work through which new systems are interpreted, negotiated, and woven into everyday practice. It is precisely these practice-based dynamics that guide our analytical focus.

To understand how autonomous systems are situated in organisational life, it is therefore not sufficient to study their technical features or design principles alone (e.g. Lund et al., 2024). It is equally important to attend to the ongoing social processes through which these technologies are made meaningful and usable. In this article, we approach the implementation of autonomous systems as something that is continuously interpreted and negotiated in everyday work – not simply rolled out from above. Drawing on Wenger's (2008) theory of Communities of Practice, we explore how learning, coordination, and shared meaning emerge around autonomous systems as part of practice-based interactions.

Wenger's framework highlights how social learning is situated in shared practices – through mutual engagement, joint enterprise, and shared repertoires (2008). Although identity is a central theme in his theory, this article sets that aspect aside in favour of a more focused analysis of the social and practical dynamics surrounding autonomous systems in mining. In doing so, we seek to better understand the organisational situatedness of these technologies not as a linear process of implementation, but as one that is deeply intertwined with the evolving practices of those who work with them.

The purpose of this study is to explore how autonomous systems are situated within organisations, and how this process is intertwined with meaning-making and social learning within communities of practice. Studies of technology adoption have also shown that adoption unfolds unevenly across a workforce, with age-related differences shaping factors such as performance expectations, social influence, and the importance of facilitating conditions (Fazi et al., 2025). This reinforces the relevance of a practice-based approach, as shared engagement and collective routines provide the setting in which such differences are negotiated and aligned in everyday work.

Our theoretical point of departure is the concept of communities of practice, which we do not employ as a comprehensive analytical framework but rather draw upon selectively to inform our understanding of the implementation process of autonomous systems in mining. The theory serves here as a sensitising concept (Bowen, 2006), binding together different empirical elements in a larger social pattern. This helps us make sense of how learning, meaning-making, and engagement evolve in relation to shared practices. We are particularly interested in how the process functions, what it is about, and what capabilities it produces.

The notion of community of practice is most commonly associated with the work of Lave and Wenger (1991), whose book Situated Learning: Legitimate Peripheral Participation marked a key shift from cognitive to social theories of learning. They introduced Community of Practice as a descriptive concept grounded in ethnographic studies of apprenticeship, focusing on informal, situated learning in practice. Around the same time, Brown and Duguid (1991) independently explored similar ideas in their work on organisational learning, helping to popularise the term within the field of management studies.

Since it was first theorised, the community of practice concept has evolved significantly. In later work (Wenger et al., 2002), it shifted from a descriptive to a more prescriptive and managerial orientation, leading to some tensions and ambiguities in how it has been theorised and applied (Cox, 2005; Gherardi, 2009). These shifts reflect not only theoretical development, but also the theory's uptake across academic disciplines and practical settings. Despite such variation, this adaptability has made the concept analytically useful in a wide range of contexts (Cox, 2005).

Much of the applied research on communities of practice has taken a different approach than ours, often focusing on how technology can be used to support knowledge sharing and learning within communities of practice, typically with an explicit interest in fostering more effective or innovative work practices. This stream of research frequently aligns with managerial goals and organisational development agendas, where communities of practice are seen as tools to improve productivity or performance. For instance, Aljuwaiber (2016) highlights how communities of practice are often implemented in business organisations as initiatives for knowledge sharing, supported by information technology and knowledge management systems. In contrast, we adopt a more interpretative stance, drawing on the theory to better understand how learning and engagement unfold in practice during a process of technological implementation.

In this study, we rely primarily on Wenger's (2008) extended treatment in Communities of Practice: Learning, Meaning, and Identity, where the focus is on practice as the locus of learning and meaning-making. Here, communities of practice are conceptualised as informal, evolving constellations of people who engage in shared activities over time, developing mutual understandings, norms, and resources. Wenger defines a community of practice along three interrelated dimensions:

1. Mutual engagement – the interactive relationships of mutual engagement that bind members into a social entity. Mutual engagement concerns how the community of practice functions.

2. Joint enterprise – the collective purpose or endeavour that members pursue and continually renegotiate. Joint enterprise concerns what the community of practice is about.

3. Shared repertoire – the set of communal resources (routines, sensibilities, artefacts, vocabulary, style, etc.) developed and used in the course of practice. Shared repertoire concerns what capabilities the community of practice has produced.

A community of practice is essentially a group of people who share an interest and improve at it by working and learning together over time. This perspective enables us to foreground the relational and situated dimensions of learning and change in the implementation process, without reducing them to either individual cognition or organisational structures.

All empirical material was collected by one author as part of a case study of the implementation of an autonomous haulage system at a Swedish open-pit mine. The data consist of 24 semi-structured interviews, complemented by field observations and fieldnotes from multiple site visits. Together, these sources provide a rich account of the practices, perspectives and relationships that emerged during the implementation.

The autonomous haulage system is a large-scale infrastructure for driverless transport in an open-pit mine. It uses self-driving trucks, each weighing 200 tonnes and carrying about 300 tonnes per load, integrated into a centralised fleet management system. This enables real-time coordination to optimise haulage routes and maintain production flow.

The machines operate in a segregated area of the mine, enclosed by barriers and access gates, forming a dedicated autonomous zone. Access is limited to machines and personnel with the required system and training. Within the zone, on-site operators manage disturbances, respond to events, and support the control room operator who oversees production remotely.

The implementation examined in this study took place in a large open-pit mine in northern Sweden employing roughly 700–1,000 people and operating continuous, high-volume production of metal ore. Annual output is substantial, involving several tens of millions of tonnes of extracted material, supported by heavy mobile equipment and tightly coordinated haulage operations. The organisation has a long history of mechanisation and incremental automation, and recent investments in digitalisation, fleet upgrades and centralised control functions have further emphasised efficiency, safety and predictable production flow. These contextual conditions shaped the environment in which the autonomous haulage system was introduced, influencing work roles, coordination demands, and the local adaptation during implementation. For instance, the former truck drivers moved into new roles, such as on-site operators, and training and competence development were emphasised for all personnel.

The interviews were collected on four separate occasions between spring 2023 and summer 2024. Participants represented diverse roles and organisational levels – from on-site and control-room operators to trainers, supervisors, section managers and programme leads. Interviews lasted 20–60 min depending on role and availability. All were audio recorded with informed consent, and participants received forms outlining the study's purpose, withdrawal rights, data handling, anonymity assurances and interviewer contact details.

An interview guide supported the interviews but was not followed rigidly; responses guided follow-up questions, while the guide ensured all core themes were covered. Twelve interviews were transcribed manually, the rest with an AI tool and then reviewed against the recordings. Most interviews were held in the “autonomous haulage building,” which housed the control room, offices and meeting rooms. A subset of on-site operator interviews took place in a stationary vehicle within the autonomous zone.

In addition to interviews, observations were crucial to data collection. The author had freedom of movement in the office building and relative freedom across the site, allowing close access to daily operations. Observations covered both activities within the autonomous zone and informal interactions in the office building. Regular presence – including shared lunches and casual conversations – enabled the author to observe everyday interactions in the emerging community of practice. These moments offered valuable insights into how work was negotiated and interpreted in real time.

Fieldnotes were written immediately after each observation and later used to support and contextualise the analysis. At times, observations were accompanied by informal interviews or spontaneous conversations, further enriching the material. The combination of interviews, observations and fieldnotes provides a nuanced, practice-oriented understanding of the implementation process, including the everyday work, negotiations and interactions that shaped the community of practice.

To analyse the data, directed content analysis was used following Hsieh and Shannon (2005). The authors began by reading the transcripts closely to obtain a sense of the whole, as one would read a novel. After this, overarching categories based on Wenger (1998, 2008) were developed, consisting of the three main aspects of communities of practice: mutual engagement, joint enterprises, and shared repertoire. Several meetings were held to define the categories and agree on their use in coding. The transcripts were divided equally and coded in NVivo over two weeks, with meetings to review progress and refine sub-categories. See Table 1 for examples of extracts, codes and categories.

Directed content analysis is well suited for qualitative studies that explore a phenomenon through a specific theoretical lens. In this study, the approach enabled the authors to use key concepts from Wenger's (2008, 1998) framework to guide the coding process and ensure alignment with the study's purpose. Because the analysis was deductive in nature, the authors were attentive to the risk of confirmation bias, namely, the tendency to privilege data that support the theoretical framework while overlooking contradictory or unexpected insights. To mitigate this, interpretations and coding decisions were continuously challenged during the analysis meetings. These joint reflections strengthened the analytical rigour by ensuring that coding was not driven by a single researcher's assumptions but was interrogated collectively. Overall, this approach enabled an analysis in which the findings are both empirically grounded in the data and theoretically informed, produced through critical engagement rather than individual judgement.

As with most qualitative studies, the generalisability of the findings is limited. The analysis reflects the specific organisational setting and practices of this particular case, and the insights should therefore be understood as context-dependent rather than universally representative. However, the strength of an in-depth qualitative design lies in its ability to illuminate the ways in which, in this case, autonomous systems are enacted, negotiated, and understood in practice. Such accounts provide insights that can inform broader theorisation and guide future research in industrial contexts.

According to Swedish law and guidelines of the Swedish Ethical Review Authority (Görman, 2023), this type of non-interventional research does not require formal ethical approval. The study did not involve sensitive personal data, physical intervention, or methods likely to cause psychological harm. Participants were informed about the purpose of the research, their right to withdraw, how data would be handled to ensure confidentiality, and anonymity. The research followed the Swedish Research Council's principles for ethical practice (Åkerman, 2024).

During the preparation of this manuscript, we used the generative AI tool ChatGPT (OpenAI GPT-4, 2025) solely to assist with linguistic refinement and phrasing – for instance, to improve fluency, grammar, and readability, and to help structure certain paragraphs. All AI-generated suggestions were carefully reviewed and revised by the authors. The authors take full responsibility for the content of the manuscript.

This section presents the results, organised according to Wenger's (2008) three dimensions of communities of practice: mutual engagement, joint enterprise, and shared repertoire. Each is illustrated with empirical examples and thematically structured into sub-sections. The findings show how the autonomous haulage system was situated in everyday work, shaped by meaning-making, social learning, and collective negotiation. Table 2 summarises the main categories and sub-categories.

A community of practice is largely defined by the mutual engagement – the interactions through which participants negotiate meaning, coordinate activities, and form social bonds (Wenger, 2008, p. 73). Practice does not exist in abstract procedures or artefacts but emerges through shared actions, conversations, and informal exchanges that establish relationships reflecting collective experience. Mutual engagement provides the structure and energy that sustain the practice over time.

A key aspect of mutual engagement in the implementation project was the widespread excitement of being part of the work with the autonomous system. On a personal level, this was described as something new and stimulating – a break from routine and an opportunity to contribute, to be involved in shaping new work practices, and to engage with what was presented as cutting-edge technology. At the same time, this novelty implied challenges, as participants were often confronted with unexpected problems that demanded creative solutions.

Crucially, this enthusiasm was not just individual but collective, providing a shared orientation that supported collaboration and reliance on each other's competences. Everyone was “being included in what matters” (Wenger, 2008, p. 74), i.e. mutually engaged and working towards the same goal. This shared commitment reinforced participants' sense of belonging and helped them manage uncertainties in the early stages. Although disagreements occurred and some had to work alone at times, the overall sense of joint commitment, fuelled by excitement about the system itself, was evident.

Given that the autonomous system was new for everyone, participants had to try out different solutions to the issues they faced during implementation. This included both technical matters, such as maximising truck efficiency, and organisational ones, such as ensuring adequate staffing. These issues were typically solved not by individuals but through joint problem-solving. Collaboration often focused on handling problems that arose in day-to-day work, especially between on-site operators and the control room:

Operators and control-room staff worked closely and were significantly dependent on each other. On-site operators were described as the control room's “eyes” or “extended arms” in the production area. If the control room made changes to the system, they relied on on-site operators' knowledge of the consequences in practice. This mutual engagement in problem-solving illustrates how competence was distributed across roles, where knowing who to ask or how to respond in specific situations was as crucial as technical knowledge. According to participants, this dynamic interdependence was one of the reasons the system worked as well as it did.

A significant shared experience was the effort to legitimise the project vis-à-vis the rest of the mine. The autonomous system and its production area differed from “normal” operations, which still relied on manual production. Participants described an emerging sense of “us and them,” as their work was sometimes viewed as problematic by those in adjacent areas. In addition to mandatory courses on the autonomous system, they took it upon themselves to actively champion the technology:

Another way of providing legitimacy involved direct engagement with the technology. One participant enthusiastically described an artificial reindeer built from cardboard, painted to resemble the real animal, and nicknamed “Reine.” It was used to test the collision avoidance system, since real reindeers sometimes wandered into the autonomous zone. Reine became something of an event within the project – placed by the coffee machine, discussed informally, and later shown in outreach material at a mining conference. The test was successful, and Reine was celebrated as a small symbol of the project:

This example illustrates how mutual engagement was enacted through relational work, humour and shared celebration. Such actions, though seemingly peripheral, were central to the community and reflect what Wenger (2008, p. 75) calls community of maintenance: the often unseen labour that sustains coherence, motivation, and meaning. The Reine episode contributed to legitimising the project both internally and externally, supporting its technical success and the ongoing negotiation of its social value and organisational relevance.

A key feature of a community of practice is negotiating a joint enterprise. Wenger (2008, p. 77) emphasises three points: it is the result of a collective process reflected in mutual engagement; it is defined by participants as they pursue it; and it is not a stated goal but created through relations of mutual accountability. In our data, participants expressed practices and activities that shaped the joint enterprise, illustrating how they made sense of their work collectively and developed an enterprise that belonged to them, not just the formal organisation.

Participants frequently described the autonomous system as improving both safety and efficiency. For operators, safety benefits included greater control over their immediate environment, such as the ability to pause activity within their “safety bubble,” thereby reducing risks of human error (e.g. tired or distracted drivers). When paused, operators could be certain that no machine would enter their space. Participants also valued that the system reduced exposure to strain injuries from long hours of truck driving. One operator stressed that monotonous tasks should be handled by machines:

Others highlighted efficiency gains, noting that autonomous trucks never take breaks and provide a smoother production flow. One participant reflected on how safety and efficiency were intertwined:

This participant first echoed management's framing of the system as a business investment, before emphasising its safety value by removing humans from hazardous tasks. Overall, participants were proud of these improvements, and regarded them as the main rationale for implementation. These reflections show that the enterprise was not simply imposed from above, but a negotiated and meaningful pursuit within the community (Wenger, 2008, 78–81).

Another joint enterprise was the belief that autonomous technology represents the future of mining, and that investment in it signals progress. The project was seen as putting the company at the forefront in Europe, enhancing its status, and attracting highly educated personnel, while also benefiting the local community:

Some participants also expressed personal motives, describing the excitement of being part of something new – “where the action is.” The project was positioned as both a technological frontier and an organisational clean slate with few established leaders or routines.

These aspirations and investments, both individual and collective, reflect what Wenger calls an indigenous enterprise (2008, p. 80): something that emerges from the community's engagement with its own conditions. Rather than simply adopting organisational goals, participants contributed to reshaping the meaning and future of work through their involvement.

Another joint enterprise was the recognition that implementing the autonomous system represented a journey of change towards becoming a more technically advanced company:

At the same time, a project leader emphasised that the main challenge was not technical, since the system was already well developed and tested elsewhere, but social:

Participants were aware of the complexity and “distance” of this journey, which was not fully understood outside the community of practice. They knew how the system functioned and had direct experience of working with it, including the training required. One control room operator described the 10-week education as far more demanding than expected:

Through their participation, community members gained insights into how the organisation would be affected that others lacked. This reflects Wenger's point that a joint enterprise is not static but constantly renegotiated, with each participant contributing through aspiration, resistance and adaptation (2008, p. 79). The project thus became a lived, shared undertaking, aligned with organisational goals but also shaped by the community's own interpretation of what mattered.

The third characteristic of a community of practice is the shared repertoire: resources for negotiating meaning developed through the joint pursuit of an enterprise (Wenger, 2008, p. 83). It includes routines, language, tools, stories, symbols, and other elements integral to practice. The repertoire enables members to make meaningful statements about the world and express identity.

In the implementation project, the repertoire was not fixed or formalised in advance but developed through practice – informed by past experience yet adapted to new contexts. Below we identify three overlapping themes that reflect this process.

Participants demonstrated confidence in the technology, grounded in an understanding of its possibilities and limitations. They knew which loading-site conditions were optimal and which created problems, from icy or bumpy roads to the layout of the autonomous zone.

Such experiential knowledge formed part of the shared repertoire – not as codified procedures but as evolving practical insights. As Wenger notes, a repertoire reflects both history and ambiguity (2008, p. 83). The ability to judge system limits and anticipate needs was rooted in situated expertise.

Participants also highlighted the extensive work required to build the autonomous zone, from constructing safety walls to laying cables and signage. Seasonal conditions were a recurrent concern:

Here, the shared repertoire also included tacit knowledge about working in a northern climate, absent from project plans and formal instructions. These embodied understandings illustrate how meaning is sustained in practice rather than through documents.

Planning, routines, and rules (formal and informal). Participants who joined the project early described how many structural factors were still undetermined. Recruitment was informal and much depended on negotiating the meaning of practice within the joint enterprise:

The early fluidity and lack of fixed routines gave participants unusual interpretative flexibility. They not only performed implementation tasks but also co-constructed the structures that later defined the autonomous system. Over time, shared language and concepts emerged; for example, “autonomous system planning” became a recognised term for long-term production coordination according to system requirements.

This repertoire was shaped by the need to move from short-term, manual production to sustained and strategically organised workflows. Participants noted that this required new ways of thinking about infrastructure, maintenance and timing, as disruptions such as cable replacements had to be carefully planned to avoid production losses.

Another key concern was ensuring the project would not be dismissed as a temporary experiment. Participants wanted the system to have what they called an organisational residence – a lasting place in the mine's routines and structure. This shows how the repertoire included not only technical routines but also interpretative and strategic resources, reflecting what would be required for the practice to endure, gain legitimacy, and spread. As Wenger (2008, pp. 83–84) notes, repertoires are resources for negotiating meaning, developed through history yet open to reinterpretation. Participants used them not only to coordinate action but to make sense of what they were doing and why it mattered.

A final aspect of the shared repertoire was balancing technological progress with organisational conditions. On the one hand, the project was presented as a long-term investment in the company's future – a way to secure recruitment, improve safety and ensure competitiveness. Yet this investment was conditional: if the system failed, the whole initiative risked being judged a failure:

On the other hand, participants faced scepticism and resistance. Some doubted the system's viability in the northern climate; others saw it as another top-down change imposed without consultation. This was compounded by organisational restructuring, widely perceived as undermining trust.

The participants found themselves caught between two forces: the progressive vision of the implementation project and the lingering doubts within the wider organisation. Navigating this tension became part of the community's repertoire – evident in how they spoke about the system, engaged with sceptics, and positioned themselves as frontiers of the new technology.

These actions reflected not only strategic thinking but also a repertoire of relational work, using interpersonal and interpretive tools to shape how the project was understood across organisational boundaries. In Wenger's terms, this illustrates how a community uses its repertoire not only to act but to situate itself within broader institutional dynamics (2008, p. 84).

The repertoire was not fixed. It emerged through ongoing engagement with the system, the organisation and other participants. Its ambiguity was a resource, enabling members to reinterpret meanings, challenge assumptions, and adapt to shifting conditions – a flexibility that kept the repertoire both powerful and alive.

Some participants developed strategies to promote acceptance of the autonomous system and ease its transition. Recognising that much resistance stemmed from fear of replacement, they sought to shift perspectives by identifying the most sceptical voices and addressing them directly – through informal conversations, demonstrations of the system, or entry-level training that granted access to the autonomous zone.

A specific strategy was to bring production leaders into the autonomous zone – not only to inform them about the technology and its potential but to give them a personal experience of how it worked.

This was done with the idea that positive impressions would spread informally within their teams. The participants viewed this type of relationship as crucial to the long-term survival of the system beyond its implementation phase.

A community of practice is not merely a group of individuals working towards a common goal – here, the implementation of a new autonomous system. Rather, participants, through daily interactions characterised by shared enthusiasm, joint problem-solving, and mutual accountability, continually negotiated and redefined their purposes. For example, the idea that the autonomous system would bring safety improvements was not an individual view from the outset but emerged through a collective process of meaning-making. Thus, joint enterprises were not imposed but created in practice through relations of mutual engagement. Their pursuit also gave rise to a shared repertoire, integral to the community's practice – expressing membership, shaping identity and strengthening participation. These findings highlight the central role of meaning-making and the situatedness of the autonomous system.

This study explored how autonomous systems are situated within organisations and how this intertwines with meaning-making and social learning in communities of practice. The findings show that successful integration depended not only on technical implementation but was deeply reliant on social processes through which participants negotiated meaning, developed shared practices and formed a collective identity around the autonomous system.

The results show that mutual engagement was crucial in shaping meaning within the community of practice. Shared enthusiasm for the autonomous system was not merely an emotional recognition of technological novelty; rather, it provided a foundation for participants to collectively invest in a common endeavour. Through daily interactions, sometimes with friction, and collaborative problem-solving, participants cultivated a sense of purpose and belonging (being included in what matters). These interactions also served as key moments of social learning, where participants exchanged knowledge, developed shared interpretations of the autonomous system and its functioning, and adapted their practices in response to emerging challenges. As Wenger (2008) suggests, mutual engagement establishes both the energy and structure of a practice, and in this case, contributed to the positioning of the autonomous system as a meaningful aspect of participants' work and professional identity.

The participant's reflections revealed that the joint enterprise surrounding the autonomous system was not simply a reproduction of organisational objectives. Rather, it was actively negotiated through collective meaning-making, particularly around improvements concerning safety and efficiency. The community articulated internally meaningful rationales for the autonomous system's value for the organisation – highlighting how it protected workers from hazardous environments and reduced monotonous labour. This internal negotiation aligns with Wenger's (2008) view that a joint enterprise is defined by the participants themselves in the pursuit of mutual accountability. In this case, meaning was generated through collective aspirations that transcended formal organisational goals.

A striking aspect of the community of practice's activity was the strategic relational work undertaken to legitimise the autonomous system within the broader organisation. Participants recognised that organisational acceptance was not automatic and therefore they actively engaged in informal outreach activities, demonstrations, and personal invitations to sceptical employees to experience the technology firsthand. This proactive engagement illustrates how meaning was negotiated beyond the immediate community boundaries, influencing wider perceptions and building social support for the autonomous system within the organisation. Such actions align with Wenger's (2008) emphasis on the relational dimensions of practice, where communities position themselves within larger organisational fields through strategic use of their shared repertoire.

The findings also underscore that meaning-making was not a static or linear process. Instead, it was continually renegotiated in response to organisational tensions, uncertainties, and shifting conditions. Participants' awareness of both the promise and fragility of the autonomous system's position within the organisation reflects the dynamic, contested nature of community enterprises (Wenger, 2008, p. 79). The community's ability to maintain a shared sense of purpose despite scepticism, organisational restructuring, and technical challenges highlights the resilience of meaning-making as a social process central to technological change. This interpretation aligns with research on ambivalent adoption in communities of practice. Schiavone's (2012) study of radio amateurs shows that communities may adopt new technologies while simultaneously maintaining older practices, resulting in hybrid forms of continuity and change. Change agents were found to play a key role in this process by working with the social and learning conditions that shape how innovations become accepted within a community. From this perspective, ambivalent responses in our case should not be seen as incomplete adoption, but rather as a durable way of negotiating technological change. In our study, such renegotiation involved continuous social learning, as the community adapted both its understanding of the technology – which, in certain aspects, started from a blank canvas – and its collective strategies for making it function appropriately.

Through mutual engagement, the negotiation of a joint enterprise, and the development of a shared repertoire, participants collectively made sense of the autonomous system, learnt from one another, defined its significance, and secured its place within the organisational landscape. The findings contribute to the understanding of how autonomous systems are not merely technological artefacts inserted into existing production systems, but socially situated phenomena whose success depends on processes of community-based meaning-making.

Importantly, the technology itself can be said to gain its meaning through the meaning-making processes within the community of practice. The function and meaning of technology are not predetermined; rather, they emerge through collective engagement and negotiation in specific social contexts. As a community of practice forms around, for instance, the implementation of a new autonomous system, meaning is not simply assigned to the technology but enacted and re-enacted through everyday practice – through problem-solving activities and shared repertoires.

In this sense, technology becomes meaningful through the ways in which it is used, interpreted, and talked about in practice. Following Wenger (2008), this process can be understood as reification – the projection of meaning onto artefacts and practices; meanings which are then perceived as having independent existences of their own. The meaning of an autonomous system, for example, may over time be reified as an apparatus protecting workers from hazardous environments, and thus also reifying a part of the community of practice itself in a congealed form. The autonomous system is no longer just a piece of highly advanced technology that a group of individuals are working with; its very existence is infused with meaning through the community of practice.

In conclusion, the function and meaning of technology are not predetermined but emerge through collective meaning-making within communities of practice. In this specific case, the community that developed around the implementation of the autonomous system played a central role in shaping how the technology was understood and made meaningful – not merely by adapting to it, but by interpreting, negotiating, and situating it in everyday practice. Thus, successful implementation cannot be understood solely as the result of skilled professionals deploying new tools; it also requires an understanding of how technology becomes meaningful within the social fabric of organisational life – what we refer to as its organisational situatedness. While this study offers a detailed analysis of how one community of practice contributed to such a transformation, further research is needed to explore how the meaning-making processes and social learning within a particular community of practice travel beyond its immediate context – for example through formal structures and/or in informal networks – and how these processes are adopted, adapted or resisted in other parts of the organisation.