Artificial intelligence (AI) literacy has recently emerged on the educational agenda raising expectations on teachers' and teacher educators' professional knowledge. This scoping review examines how the scientific literature conceptualises AI literacy in relation to teachers' different forms of professional knowledge relevant for Teacher Education (TE). The search strategy included papers and proceedings from 2000 to 2023 related to AI literacy and TE as well as the intersection of AI and teaching. Thirty-four papers were included in the analysis. The Aristotelian concepts episteme (theoretical-scientific knowledge), techne (practical-productive knowledge), and phronesis (professional judgement) were used as a lens to capture implicit and explicit dimensions of teachers' professional knowledge. Results indicate that AI literacy is a globally emerging research topic in education but almost absent in the context of TE. The literature covers many different topics and draws on different methodological approaches. Computer science and exploratory teaching approaches influence the type of epistemic, practical, and ethical knowledge. Currently, teachers' professional knowledge is not broadly addressed or captured in the research. Questions of ethics are predominantly addressed as a matter of understanding technical configurations of data-driven AI technologies. Teachers' practical knowledge tends to translate into the adoption of digital resources for teaching about AI or the integration of AI EdTech into teaching. By identifying several research gaps, particularly concerning teachers' practical and ethical knowledge, this paper adds to a more comprehensive understanding of AI literacy in teaching and can contribute to a more wellinformed AI literacy education in TE as well as laying the ground for future research related to teachers' professional knowledge.

This article explores the complex challenges of co-designing an AI- and learning analytics (LA)-integrated learning management system (LMS). While co-design has been proposed as a human-centred design approach for scaling AI and LA adoption, our understanding of how these design processes play out in real-life settings remains limited. This study is based on ethnographic fieldwork in primary and secondary schools and employs a relational materialist approach to trace, visualise, and analyse the increasingly complex and transformative relations between a growing number of actors. The findings shed light on the intricate ecosystem in which AI and LA are being introduced and on the marketisation of K-12 education. Instead of following a rational and sequential approach that can be easily executed, the co-design process emerged as a series of events, shifting from solely generating ideas with teachers to integrating and commercialising the LMS into a school market with an already high prevalence of educational technology (EdTech). AI and LA in education, co-design and data-driven schooling served as negotiating ideas, boundary objects, which maintained connectivity between actors, despite limited AI and LA implementation and the development of a stand-alone app. Even though teachers and students were actively involved in the design decisions, the co-design process did not lead to extensive adoption of the LMS nor did it sufficiently address the ethical issues related to the unrestricted collection of student data.

Machine learning and other artificial intelligence (AI) technologies are predicted to play a transformative role in primary education, where these technologies for automation and personalization are now being introduced to classroom instruction. This article explores the rationales and practices by which machine learning and AI are emerging in schools. We report on ethnographic fieldwork in Sweden, where a machine learning teaching aid in mathematics, the AI Engine, was tried out by 22 teachers and more than 250 primary education students. By adopting an Actor-Network Theory approach, the analysis focuses on the interactions within the network of heterogeneous actors bound by the AI Engine as an obligatory passage point. The findings show how the actions and accounts emerging within the complex ecosystem of human actors compensate for the unexpected and undesirable algorithmic decisions of the AI Engine. We discuss expectations about AI in education, contradictions in how the AI Engine worked and uncertainties about how machine learning algorithms ‘learn’ and predict. These factors contribute to our understanding of the potential of automation and personalisation—a process that requires continued re-negotiations. The findings are presented in the form of a fictional play in two acts, an ethnodrama. The ethnodrama highlights controversies in the use of AI in education, such as the lack of transparency in algorithmic decision-making—and how this can play out in real-life learning contexts. The findings of this study contribute to a better understanding of AI in primary education.

This paper provides a critical examination of the domain of artificial intelligence (AI) in education, with a focus on the expectations and practical implications accompanying its integration into teaching. The expectations have been propelled by two interconnected concepts: (1) the potential for AI to automate pedagogical processes, replacing teachers in certain scenarios; and (2) the notion that teachers’ insights can be augmented through AI-based analysis. Drawing on two ethnographic studies in Swedish primary and secondary schools, this paper explores the enactments of pupils, teachers and two AI-based educational technologies. The aim is to demonstrate how automation and augmentation can emerge in teachers’ practice. Utilizing inspiration from a relational epistemological problematisation of socio-technical phenomena, the paper demonstrates how rather than automation and augmentation, AI in education is an act of symmation in which automation and augmentation is co-produced by the technology and teachers’ different hidden work, in this paper conceptualised as adaptations, experimentations, compensations and confirmations. The paper suggests that the study of symmation in relation to the teaching profession can be productive in further exploring the yet limited understanding of AI in educational practice.

This paper explores the implementation of a wellbeing app in a Swedish Upper Secondary School. The aim is to understand how ideas of data driven school improvement underpinned by promises of artificial intelligence (AI) and learning analytics (LA) change the work of teachers. The study draws on video-ethnography from 17 meetings between five teachers/form tutors. The produced data is analysed using actor-network theory to focus on the various stages of the implementation process and the interactions between the learning analytics dashboard (LAD) and the teachers. To capture the complexity of the data, the empirical material is presented through cartoon-inspired illustrations grounded in a Thinking through Cartoons methodology. Findings show how teachers took on new roles and responsibilities in relation to the wellbeing app, most notably the role of collecting data from students. Teachers came to act as data analysts which imposed constant negotiations and uncertainties. To address the declining engagement of students over time, a student-facing LAD was intro-duced. The teachers shifted their focus to motivate students to engage with their own data in different ways. Despite no improvements in students’ response rates teachers remained committed to the app, trusting that new AI and LA function-alities would compensate unsatisfactory outcomes. In conclusion, instead of im-proving teachers’ capacity to identify at-risk students, the wellbeing app in-creased teachers’ workload and led to different dilemmas related to teacher-stu-dent relations and teachers’ professional judgement.

This method-focused article continues the discussion of how methodological challenges in research with children with disabilities can be overcome. By examining the difficulties in introducing walk-and-talk conversations in an ethnographic study with young children (3-11 years) with disabilities, it asks what doing research with means and on whose terms research should take place. The article argues for a shift in the understanding of doing with, from a researcher-controlled research encounter to the focusing of childrens own activities. This ensures the participation of all children in research, by recognizing and listening to the many different ways in which children communicate.

Research has established that children with disabilities face barriers to mainstream leisure. But what happens when leisure is specifically organised for children with disabilities? This article builds on an ethnographic study at a 'separate', i.e. disability-specific, leisure space for young children (3-11 years) with disabilities, such as ADHD and autism. It asks: What is made possible by separate leisure and are there potential limits? The analysis shows that separate leisure plays an important role as a site for leisure outside children's homes. Whereas children with disabilities are typically the exception in mainstream leisure contexts, separate leisure positions their needs and desires as the starting point for designing and organising leisure. Moreover, separate leisure can act as a 'leisure node' and a context for leisure without expectations about children's performances. The study highlights the need to challenge societal perceptions of leisure, particularly the belief that leisure is best enjoyed with others, to better accommodate all children's leisure. It suggests that separate leisure holds potential to advance theory and practice by illustrating how leisure can be understood and practiced in a context where ableist and adult-centred assumptions are not at the forefront.

Separate leisure spaces play an important part in children with disabilities everyday geographies, though little is known about how they are designed and organised or how children use them. This article contributes to the field of disabled childrens geographies (cf. Ryan, Sara. 2005. "People Dont do odd, do They? Mothers Making Sense of the Reactions of Others Towards Their Learning Disabled Children in Public Places." Childrens Geographies 3 (3): 291-305) by building on a three-year ethnographic study that explores a separate leisure space in Sweden for children (3-11 years) with disabilities such as ADHD and autism. The current article focuses on the calm room within the facility, aimed at providing space for preventing and handling children acting out. By analysing this particular room, the study illuminates the ideas and assumptions about the children that went into designing the separate leisure space. Two dimensions of the room are analysed: (i) the design of the room, and (ii) childrens uses of the room.The analysis demonstrates that children used the room for their own purposes, for example for resting, socialising or playing. When childrens uses of the room conflicted with what designers had planned for, tensions arose between the ideas about childrens needs that informed the design of the room and what the children needed or wanted during their visits. This demonstrates the importance of not having too rigid ideas about children with disabilities needs when planning and designing separate leisure spaces. It is suggested that one way of ensuring that childrens actual needs and desires are considered, rather than those assumed or imagined by adult designers, is by finding ways to include children in the design and planning processes.

Inflammatory diseases, including psoriasis, are characterized by changes in redox regulation. The MTH1 prevents the incorporation of oxidized nucleotides during DNA replication. Using MTH1 small-molecule inhibitors, we found induced apoptosis through 8-oxodeoxyguanosine triphosphate accumulation and DNA double-strand breaks after oxidative stress in normal and malignant keratinocytes. In psoriasis, we detected increased MTH1 expression in lesional skin and PBMCs compared with that in the controls. Using the imiquimod psoriasis mouse model, we found that MTH1 inhibition diminished psoriatic histological characteristics and normalized the levels of neutrophils and T cells in the skin and skin-draining lymph nodes. The inhibition abolished the expression of T helper type 17-associated cytokines in the skin, which was in line with decreased levels of IL-17-producing gd T cells in lymph nodes. In human keratinocytes, MTH1 inhibition prevented the upregulation of IL-17-downstream genes, which was independent of ROS-induced apoptosis. In conclusion, our data support MTH1 inhibition using small molecules suitable for topical application as a promising therapeutic approach to psoriasis.

Psoriasis is an IL-23/Th17-mediated skin disorder with a strong genetic predisposition. The impact of its susceptibility gene nitric oxide synthase 2 (NOS2) remains unknown. Here, we demonstrate strong NOS2 mRNA expression in psoriatic epidermis, an effect that is IL-17 dependent. However, its complete translation to protein is prevented by the IL-17-induced miR-31 implying marginally upregulated NO levels in psoriatic skin. We demonstrate that lower levels of NO, as opposed to higher levels, increase keratinocyte proliferation and mediate IL-17 downstream effects. We hypothesized that the psoriatic phenotype may be alleviated by either eliminating or increasing cellular NO levels. In fact, using the imiquimod psoriasis mouse model, we found a profound impact on the psoriatic inflammation in both IMQ-treated NOS2 KO mice and wild-type mice treated with IMQ and the NO-releasing berdazimer gel. In conclusion, we demonstrate that IL-17 induces NOS2 and fine-tunes its translation towards a window of proinflammatory and hyperproliferative effects and identify NO donor therapy as a new treatment modality for psoriasis.

Several European railway traffic networks experience high capacity consumption during large parts of the day resulting in delay-sensitive traffic system with insufficient robustness. One fundamental challenge is therefore to assess the robustness and find strategies to decrease the sensitivity to disruptions. Accurate robustness measures are needed to determine if a timetable is sufficiently robust and suggest where improvements should be made.

Existing robustness measures are useful when comparing different timetables with respect to robustness. They are, however, not as useful for suggesting precisely where and how robustness should be increased. In this paper, we propose a new robustness measure that incorporates the concept of critical points. This concept can be used in the practical timetabling process to find weaknesses in a timetable and to provide suggestions for improvements. In order to quantitatively assess how crucial a critical point may be, we have defined the measure Robustness in Critical Points (RCP). In this paper, we present results from an experimental study where a benchmark of several measures as well as RCP has been done. The results demonstrate the relevance of the concept of critical points and RCP, and how it contributes to the set of already defined robustness measures.

Maintaining high on-time performance and at the same time having high capacity utilization is a challenge for several railway traffic systems and especially those with heterogeneous traffic. With high capacity utilization the system is sensitive to disturbances and delays could easily propagate in the network. One way to handle this problem is to create more robust timetables; timetables that can absorb delays and prevent them from propagating. This paper presents an optimization approach to reduce the propagation of delays by introducing a more efficient margin time allocation in the timetable. A Mixed Integer Linear Programming (MILP) model is proposed, in which the existing margin time is re-allocated to increase the robustness of an existing timetable. The model re-allocates both runtime and headway margin time to increase the robustness at specific delay sensitive points in a timetable, a suitable approach for double-track lines with dense heterogeneous traffic. We illustrate the applicability of the approach in a real-world case, where an initial timetable is modified into new timetables with increased robustness. These new timetables are then evaluated and compared to the initial timetable. We evaluate how the re-allocation of margin time affects the timetable structure and the timetable's capability to handle disturbances by exposing it to some minor initial disturbances in the range of 5–10 min. The results show that it is possible to reduce the delays by re-allocating the existing margin time. For example, the total delay at end station decreases with 10% in our real-world example.

One method to increase the quality of railway traffic flow is to construct a more robust timetable in which trains are able both to recover from delays and the delays are prevented from propagating. Previous research results show that the indicator Robustness in Critical Points (RCP) can be used to increase timetable robustness. In this paper we present the use of a method for RCP optimization: how it can be assessed ex-post via microscopic simulation. From the evaluation we learn more about how increased RCP values influence a timetable's performance. The aim is to understand more about RCP increase at a localised level within a timetable in terms of effects to the pairs of trains that are part of the indicator. We present a case study where an initial timetable and a timetable with increased RCP values are evaluated. The ex-post evaluation includes the quantification of measures concerning train-borne delay and robustness of operations, as well as measures capturing the subsequent quality of service experienced by passengers to assess the broader effects of improved robustness. The result shows that it is necessary to use several key performance indicators (KPIs) to evaluate the effects of an RCP increase. The robustness increases at a localised level, but the results also indicate that there is a need to analyse the relationship between ex-post measures and RCP further, to improve the method used to increase RCP and thus its overall effect on timetable robustness.

Due to high demand and capacity consumption, railway timetables are often sensitive to disturbances. To maintain punctual operations, it is important that timetables are robust, and methods are needed that make them robust without consuming too much capacity. In this paper, we demonstrate how a policy change in the form of new timetable planning rules can be used to achieve more robust timetables. We present the use of the rules in a real-world case from 2019, when our rules were applied for the Swedish Southern mainline. In this paper, we describe how a new policy for scheduling trains can be applied, and we discuss implications observed when going from research to practice. We also describe how the proposed rules affect train paths and runtimes. The outcome of the rules is measured in a comprehensive evaluation of the traffic performance based on empirical operational data. The results from this study show that practical knowledge is necessary when developing a policy, as well as when developing a timetabling model. Insights, given to us by experienced timetable planners, can be used to enhance optimisation models and make the models more applicable in the real world. The main contribution of this paper is to show that it is possible to increase timetable robustness with a minor policy change based on previously presented research results. Even with relatively small timetable modifications, we can learn from the operational data that the new rules had the intended effect and that overall punctuality can be increased.

When several trains are planned to use the same infrastructure resource, there is always a risk for spreading of delays, which can be hard to recover from. It is a challenge for the Infrastructure Manager to make timetables that accommodate as much traffic as possible, without causing bad on-time performance. Timetable planners are in need of quantitative indicators to assess timetable robustness and accurate methods for how to make the timetable more robust.

In this paper we assess the robustness for single-track lines with non-periodic timetables. At single-track lines, trains use the line for running in both directions and the trains can only pass or overtake each other at passing loops. This makes the system more sensitive for delays. In this paper we present a robustness indicator which captures the dependencies between trains at a single-track line. The indicator can be used to illustrate weaknesses in a timetable and also to indicate where and how to insert more robustness. In a simulation study, we show that it is possible to improve the performance by making small timetable adjustments according the indicator, without increasing runtimes or capacity utilization.

Studsvik, a Swedish nuclear research facility, has been releasing aquatic radioactive discharges in the Baltic Sea, through the bay Tva center dot ren, since 1959. The permissible discharge levels are regulated by the Swedish Radiation Safety Authority (SSM) but only information about 60Co, 137Cs, 152Eu, total alpha and beta activities were re-ported up to 2002. Since then, the reports cover most a more comprehensive set of radionuclides. In this context, the seabed can be utilized as a chronological archive to investigate historical Studsvik releases. To this end, 23 sediment cores covering the whole area of the bay were studied and 5 of them were dated using 210Pb-dating methods. Since the discharges from Studsvik contain both plutonium and caesium, neither can be used to validate the 210Pb-dating method. Instead, stable lead with maximum deposition, known to be dated to 1970, was used. Cobalt-60, 137Cs, and 152Eu depth distributions were studied from the dated sediment cores and compared with reported levels of aquatic discharges. The expected sediment 137Cs-peak from the fallout from the Chornobyl accident was however smeared out, indicating an ongoing inflow of 137Cs with the Baltic seawater. Our findings show that reported releases of aquatic discharges of 60Co and 152Eu are consistent with measured sediment activity distribution. The sediments from the deepest parts of Tva center dot ren, with intact chronology and with a high time resolution, are ideal for investigating historical nuclear discharges and can be a tool to investigate unreported radionuclide releases. Dated sediment can in this way be a tool for nuclear safeguards to evaluate past and present activities in nuclear facilities.

Plutonium distribution was studied in an undisturbed sediment core sampled from the Tvaren bay in the vicinity of the Studsvik nuclear facility in Sweden. The complete analysis, including minor isotopes, of the Pu isotope composition (238Pu, 239Pu, oPu, 241Pu, 242Pu, and 244Pu) allowed us to establish the Pu origin in this area of the Baltic Sea and to reconstruct the Studsvik aquatic release history. The results show highly enriched 239Pu, probably originating from the Swedish nuclear program in the 1960s and 1970s and the handling of high burn-up nuclear fuel in the later years. In addition, the 244Pu/239Pu atomic ratio for the global fallout period between 1958 and 1965 is suggested to be (7.94 +/- 0.31)center dot 10-5. In the bottom layer of the sediment, dated 1953-1957, we detected a higher average 244Pu/239Pu ratio of (1.51 +/- 0.11)center dot 10-4, indicating the possible impact of the first US thermonuclear tests (1952-1958).

Ten percent of the Swedish population over the age of sixteen can be counted as digitally excluded. Many of them turn to public libraries for support with e.g. public e-services. As street-level bureaucrats, library staff can support citizens’ digital inclusion within the scope of their assignment and institutional framework. However, they face dilemmas when they encounter citizens with needs that fall partly outside this framework.

According to a classic work of Weber, bureaucrats can be seen as locked in an iron cage of rules and norms. It could be argued that digitalization is changing the conditions in bureaucracies even further. As a complement to Weber’s iron cage, Peeters and Widlak described the disciplining logic of digital information architecture as a form of digital cage that can exclude citizens and frame the discretion of street-level bureaucrats.

This article elaborates on the concept digital cage, built on the iron cage concept, to analyse how street-level bureaucrats cope with discretionary boundaries in their work for citizens’ digital inclusion. Based on an analysis of interviews with library staff at local centres for digital inclusion, policy documents and interviews with local politicians in charge of these centres, the study shows the relevance of the digital cage concept and that the use of new technologies changes and partly limits street-level discretion. However, it also shows that politicians allow library staff a considerable degree of freedom in their work, and that the centres’ focus on face-to-face meetings and entrepreneurial solutions is an asset for digital inclusion.

Digital technology is frequently used in the delivery of public services. Since not everyone has the skills or digital access required to use such digital services, street-level bureaucrats must find new ways to support citizens to be able to enhance digital inclusion. In most public settings, these new practices of digital support need to be developed in parallel with the street-level bureaucrats' ordinary assignments. This article analyses the development of street-level bureaucrats' entrepreneurial and innovative practices in promoting digital inclusion in two spheres: conventional welfare service provision and in a project-based welfare service provision initiative explicitly intended to enhance digital inclusion. The article builds on qualitative case studies in municipalities in Sweden and demonstrates that there are more openings for innovative, entrepreneurial solutions among street-level bureaucrats in the more project-based setting than in the conventional welfare services. This implies that context matters for street-level entrepreneurial approaches.

This paper maps and unpacks the discursive construction of digital citizenship within the institution of public libraries in Sweden, situating it within the broader context of the nation’s digital government. By adopting a discursive approach, the paper responds to the scholarly call for nuanced theoretical insights into the evolving digital citizenship field and reflects upon how an analysis on digital citizenship can further our understanding of digital government. Data sources are the Swedish National Digitalization Strategy, the Swedish Library Act, and library staff interviews.

Findings show active citizenship as a central idea in the discourse, placing high demands on individuals. In contrast to previous research, this study also finds ideas of a more inclusive citizenship. Finally, the findings show conflicts and ongoing negotiation of what digital citizenship, and thus digital citizens, ought to be, and point towards a need of deepening and questioning our understanding of digital government as inclusive.  

As artificial intelligence (AI) is increasingly integrated into the context of complex products and systems (CoPS), making complex systems more intelligent, this article explores the consequences and implications for engineering management in emerging complex intelligent systems (CoIS). Based on five engineering management aspects, including design objectives, system boundaries, architecting and modeling, predictability and emergence, and learning and adaptation, a case study representing future CoIS illustrates how these five aspects, as well as their relationship to criticality and generativity, emerge as AI becomes an integrated part of the system. The findings imply that a future combined perspective on allowing generativity and maintaining or enhancing criticality is necessary, and notably, the results suggest that the understanding of system integrators and CoPS management partly fundamentally alters and partly is complemented with the emergence of CoIS. CoIS puts learning and adaptation characteristics in the foreground, i.e., CoIS are associated with increasingly generative design objectives, fluid system boundaries, new architecting and modeling approaches, and challenges predictability. The notion of bounded generativity is suggested to emphasize the combination of generativity and criticality as a direction for transforming engineering management in CoPS contexts and demands new approaches for designing future CoIS and safeguard its important societal functions.

Rapid developments in artificial intelligence (AI) are driving the evolution of complex products and systems (CoPS) into complex intelligent systems (CoIS). The introduction of AI implies generativity and increasingly fluid boundaries in such systems and presents challenges for organizations to control and manage systems that are safety critical. Building on a case study representing future CoIS, this paper explores fluid boundaries in CoIS, including approaches for navigating system criticality and generativity. The findings point to the relationship between fluid boundaries and a stable organizational and system core, along with a shared core mission. Together, they serve as a platform that enables both contributions from various constituent systems and dynamic reconfigurations of the overall system-of-systems (SoS). System criticality and generativity are navigated through setting bounds to generativity by checks and balances involving both human and AI, including safety requirements for constituent systems and overall human oversight. Such an approach extends beyond traditional system integration activities and alters the role of CoIS integrators.

As artificial intelligence (AI) is becoming part of complex systems and critical infrastructures, perspectives on resilience may need to be revisited. This paper focuses on the challenges and approaches in engineering design for achieving resilience in complex and increasingly intelligent systems (CoIS). Building on a case study of a system situated in the context of search and rescue (SAR) operations at sea as well as scenarios of SAR operations supported by AI solutions, it outlines challenges for organisational and engineering design in contexts where flexibility, adaptability, and high reliability are important. The findings point at resilience as a system property, made up of the constituent systems, their interaction and coordination in a system-of-systems framework. AI and autonomy in CoIS represent potentially a double-edged sword; while AI and autonomy contribute to system capabilities and resilience, they can also introduce limitations in terms of, for instance, confined operational envelopes. Achieving resilience in CoIS thus requires a holistic approach that considers constituent systems as well as their interplay, organisational factors, and the judicious balance of AI and human-based solutions.

With the emergence of artificial intelligence (AI), data-driven innovation has become a new paradigm in innovation management. As these solutions are increasingly integrated into complex, safety-critical systems, data becomes crucial in these contexts. However, safety-critical applications require high-quality and verified data in areas where sometimes only sparse data is available. This paper explores the role of data in these contexts and sheds light on data and innovation in complex intelligent systems (CoIS). The findings reveal that a lack of available data may hinder the innovation process and that the necessary training data must be generated through the innovation itself. The results thus show that the emerging data-driven innovation paradigm needs to be complemented with a focus on understanding innovation-driven data generation in the context of systems with safety-critical applications.

The rapid growth, large volume, and easy access to publicly available netnographicdata have increased the need for understanding its potential in leveraging innovation.At the heart of quality management is a focus on customers and their needs, drivingthe innovation of products and services. In this study, we systematically comparenetnography and focus groups as research methods regarding their ability to capturevarious types of customer needs. This study makes four valuable contributions toquality management theory: (1) netnography and focus groups captures similarcustomer needs, (2) netnography provides clear need statements, which are rich andoverrepresented compared to focus groups, (3) both netnography and focus groupsare vulnerable to groupthink, where customers prioritize consensus; however, thiseffect is more evident in netnography, (4) netnography captures more needstatements about current product and service usage, while focus groups focus moreon past product and service usage

Purpose – The integration of industry 4.0 has become a priority for many organizations. However, not allorganizations are suitable and capable of implementing industry 4.0 because it requires a dynamic andflexible implementation strategy. The implementation of industry 4.0 often involves overcoming severaltensions between internal and external stakeholders. This paper aims to explore the paradoxical tensions thatarise for health-care organizations when integrating industry 4.0. Moreover, it discusses how a paradox lenscan support the conceptualization and proposes techniques for handling tensions during the integration ofindustry 4.0.

Design/methodology/approach – This qualitative and in-depth study draws upon 32 semi-structuredinterviews. The empirical case concerns how two health-care organizations handle paradoxical tensionsduring the integration of industry 4.0.

Findings – The exploration resulted in six recurring technology tensions: technology invention(modularized design vs. flexible design), technology collaboration (automation vs. human augmentation),technology-driven patient experience (control vs. autonomy), technology uncertainty (short-termexperimentation vs. long-term planning), technology invention and diffusion through collaborative effortsamong stakeholders (selective vs. intensive collaboration) and technological innovation (market maintenancevs. disruption).

Originality/value – A paradox theory-informed conceptual model is proposed for how to handle tensionsduring the integration of industry 4.0. To the best of the authors’ knowledge, this is the first paper to introduceparadox theory for quality management, including lean and Six Sigma.