Automation–Augmentation Entanglement: Self-Service Technology as Driver for Change

VOLUME XX, 2017 1 Date of publication xxxx 00, 0000, date of current version xxxx 00, 0000. Digital Object Identifier 10.1109/ACCESS.2024.Doi Number Automation–Augmentation Entanglement: Self-Service Technology as Driver for Change Sophie Altrock1,2, Anne-Laure Mention1,3,4,5, and Tor Helge Aas2,6 1RMIT University, School of Management, Melbourne, 3000 VIC, Australia 2University of Agder, School of Business and Law, 4630 Kristiansand, Norway 3Tampere University, 33100 Tampere, Finland 4Singapore University of Social Sciences, 599494 Singapore 5INESC TEC Portugal, 4200-465 Porto, Portugal 6Kristiania University College, 0107 Oslo, Norway Corresponding author: Sophie Altrock (e-mail: [email protected]). This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 956745. Results reflect the author’s view only. The European Commission is not responsible for any use that may be made of the information it contains. ABSTRACT Although the automation of work is not new, advancements in artificial intelligence, worker shortages, and the Covid-19 pandemic have accelerated recent debates. Especially in service sectors, traditional service delivery is gradually merging with machine components, requiring organizations to rethink models of service delivery. Since the adoption of self-service technology (SST) grows, users benefit from improved access and self-management without human touchpoint, yet the impact of SST on workers requires further exploration. While technology-enabled workplace transformation is well researched, unified frameworks and strategies are missing to understand the interplay of augmentation and automation as SST becomes a ubiquitous component of service. Through an integrative review we explore most cited and most recent literature on automation, augmentation, and SST. In total, we review 41 articles. This review provides a status quo on the topics automation, augmentation, and SST. Providing a combined lens, five emerging themes are discovered for future research: Social (A)symmetries, Competing Intelligences, Redistributed Responsibilities and New Skills, Merging Spaces and New Ecosystems, Governance and Ethics. This article provides an interdisciplinary review of service automation, extends previous scholarly work, and encourages new pathways for hybrid service delivery. Furthermore, embedded in the concept of a human-centered Industry 5.0, we suggest automation–augmentation entanglement as holistic framework reflecting the interplay of human and technology in the workplace. INDEX TERMS service automation; industry 5.0; augmented work; self-service technology; artificial intelligence; integrative review I. INTRODUCTION Although automation has been around for decades, the concept and our understanding of it is undergoing a gradual shift. While digital technologies reengineer various tasks whereby moving task responsibility from human worker to machine, increasingly more tasks are affected [1]. With recent developments in artificial intelligence (AI) the possibilities to automate are growing, driving the debate of how automated processes will merge with and alongside human workers [2]. Since the first industrial revolution, the role of technology and workers has been shifting, gradually evolving into industry 5.0. As suggested by the European Union [3], progressing from industry 4.0, the new concept aims for sustainable, human-centric, and resilient industries that maximize strengths of both, technology and human, “by understanding where each excels” [4]. On the social impact level, industry 5.0 offers values such as job satisfaction, upskilling and reskilling, human agency, or improved customer experience [5]. By extension, automation today does not simply express the replacement of human workers but rather the enhancement or augmentation of human skills. Therefore, scholars have This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2025.3620137 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ VOLUME XX, 2017 7 noted that automation and augmentation exist at the same time and in the same space [6]. As noted by Raisch & Krakowski [6, p. 204], “we have to acknowledge that these human and machine agents do not simply coexist in separate worlds (working on separate tasks), but are interdependent (interacting on the same or closely related tasks)”. Among the affected sectors, the professional services sector in particular questions the role of human workers where human interaction is often fundamental, and expertise is acquired over many years. Recently, advancements in AI have “brought professional services into the automation arena” [7, p. 124]. As such, increasingly smart technologies can take over certain tasks originally undertaken by workers or provide services directly to customers without the need of human intervention. For instance, robo-advisors have opened up online investment services through weband app-interfaces without the need for human financial advisors, automating the process of investing. However, scholars have started noting the need for hybrid models of service delivery, including human and robo-advisors along the process to account for technological shortcomings [8]. As the technologies advance, individuals become empowered and turn to online services without human touch. With self-service technologies (SSTs) such as the robo-advisor, self-checkouts in supermarkets, and symptom checkers for medical diagnoses, customers have become co-producers of service delivery, often making additional service interactions from human to human redundant [9]. Thus, while SST becomes available and customers turn into co-creators, the effects on workers and the shift to such models have to be understood—an area that has been receiving growing attention [10], [11]. Although co-creation and co-production are two established concepts in the literature [12], value generation in service interactions is further transforming through SST adoption and increasing recognition of augmentation. As such, co-creation emphasizes “joint effort and collaboration between producer and the consumer”, underpinning “reciprocity and mutuality” in such interactions [12, p. 11]. However, the role of SST in that equation has not yet received much attention although “[t]he provision of service involves both employees and customers and can be provided by humans and/or machines” [2, p. 156]. While much research has been looking at service automation, self-service, and the changing nature of service encounters, few scholars have explored SST in the context of automation and augmentation, calling for a better understanding of service augmentation [7], [13], [14]. In particular an interdisciplinary lens is missing as research often looks at either technology adoption on the consumer side or at service aspects from the organizational side, missing out on understanding how the two are connected. With the growing acknowledgment of an interdependence between automation and augmentation, a more holistic view on automation would allow us to understand automation not only from a “machine replacement” perspective but will open up a comprehensive “human augmentation” perspective that industry 5.0 calls for [3]. While we know that the process of introducing digital technology into organizations requires a holistic view on the micro-, meso-, and macro levels, e.g., [15], [16], we have yet to develop a holistic understanding of automation. A timely debate requires the acknowledgment of a significant lack of skills needed to advance into industry 5.0 [17]. Further challenges are represented by staff shortages across service industries where hybrid service provision and the mobilization of individuals on the consumer side is crucial in order to even sustain those services – especially in sectors such as finance, healthcare, education, or social services. Therefore, with this paper we aim to bridge the literature on automation, augmentation and SST to understand their interdependencies and their mutual impact on service users and service workers. Presented as varying archetypes by De Keyser et al. [10], we zoom in on the role of SST for augmented service encounters. As “[f]irms should design human–machine integrated service strategies […]” [2, p. 167], this paper aims to contribute to the further development of human-machine services for competitive edge, but, more so, for worker wellbeing and quality services. We do so by answering the following RQs: • RQ1: What do we know about automation, augmentation and SST and how SST affects service workers? • RQ2: How do automation, augmentation, and SST interact and what are emerging themes for future research? II. METHOD A. RESEARCH AIM The aim of this paper is to provide a holistic understanding of the effects of SST on automation and augmentation in service industries as customers become increasingly engaged and informed, challenging the status quo of professionals across sectors. Though each of the topics alone have been studied in depth in their respective disciplines, what is currently missing is a contextualized approach that combines different viewpoints and provides a connected and integrative synthesis of the literature. In particular, as we see a significant growth in digital SST adoption (through AI and robotic applications specifically) and a rising recognition of augmentation, pinning the status quo of the debates helps to provide strategies moving forward. Therefore, we choose an integrative literature review, to draw on multi-disciplinary research in order to examine the role and interconnectedness of automation, This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2025.3620137 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ VOLUME XX, 2017 7 augmentation, and SST as they relate to knowledge workers. As such, we seek to consolidate and synthesize current research into an extended framework for service automation to guide further research, practice, and policy. We select the integrative review in particular, as we believe that the topic “would benefit from a holistic conceptualization and synthesis of the literature” [18, p. 410] and the integrative review is shown to be “a powerful tool for studying organizational phenomena” [19, p. 187]. With the emergent empowerment of customers through the means of automation and its growing effects on workers through augmentation, we choose the integrative review to review, critique, and synthesize representative literature, and to reconceptualize and extend our understanding of automation [18, p. 411]. In this paper we build onto the concepts established by Raisch & Krakowski [6] and Sampson [7], [20] and suggest that SST through customer automation can advance worker augmentation—grounded in furthering the concept of industry 5.0. As the literature on these thematic areas remains dispersed, we provide an overview of these areas and how those can be consolidated. The goal of our selection process, as described below, was therefore not to be exhaustive but rather to engage more deeply with widely discussed and recent research to inform future research and theory development. B. LITERATURE SELECTION To select relevant literature for the review, a title, abstract, and keyword search was conducted using the Scopus database, considered one of the largest databases. For that purpose, a total of eight searches was undertaken using the search terms automat*, augment*, and self-servi* each with AND “service” AND “digital” as well as the five search terms altogether, automat* AND augment*, and in combination with self-servi* and all three search terms together. Further, each search was narrowed down by focusing on English journal and review articles in three subject areas; (1) the social-sciences to understand the sociotechnical aspects of human-machine interaction, (2) business, to explore the management of technology adoption in organizations, and (3) economics, as much of the automation literature originates from the economist’s perspective [21], [22]. A preview of the automation literature reveals some often-cited works driving the debate. Thus, further narrowing down the literature and focusing the debate, highly cited articles were selected for review [23], [24], selecting the 10 most-cited articles for each search as well as the 10 most-cited articles for 2024 with respect to the topical dynamic and pace. In addition, the 10 most recent articles were selected to include the most recent developments and discussions in the field regardless of citation count which may not suffice as a standalone selection criterium. Combining all search results led to a collection of 189 articles from which 59 duplicates were removed. From 130 papers another 71 were excluded that were, upon first inspection, not addressing the sociotechnical effects of automation, augmentation or self-service such as papers with an engineering perspective. 59 papers were screened further, resulting in the exclusion of 18 additional papers as their research focus did not fit the research scope and question. This finally resulted in 41 papers for in-depth review. The results from this process can be found in Table I. TABLE I SEARCH TERMS AND NUMBER OF LITERATURE RESULTS Search terms Total results 10 most highly cited 10 most highly cited in 2024 10 most recent augment* AND service AND digital 337 10 10 10 automat* AND service AND digital 1113 10 10 10 self-serv* AND service AND digital 105 10 10 10 self-serv* AND augment* 25 10 2 10 self-serv* AND automat* 152 10 10 10 automat* AND augment* 1171 10 10 10 automat* AND augment* AND selfserv* 6 6 1 6 automat* AND augment* AND selfserv* AND service AND digital 2 2 0 2 Total 2911 68 53 68 SELECTION ROUND I Highly cited + recent 189 After duplicates removed 130 SELECTION ROUND II After initial scoping 59 SELECTION ROUND III Final sample after review 41 As the Scopus database search shows, less than 3000 research and review articles can be found in the selected research disciplines, addressing some aspects of automation, augmentation, self-service, service, and/or digital. Most of those papers mention automation and augmentation together, followed by papers addressing automation, service, and digital at once. As Figure 1 shows, papers addressing automation and augmentation across all disciplines and publication types have This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2025.3620137 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ VOLUME XX, 2017 7 significantly increased in the last 8 years. Most of these, a combined 52,7%, are in the Computer Sciences and Engineering fields, 3,8% in the Social Sciences and even less in the fields of Business and Economics. Although SST has seen a similar but less significant increase since the 2000’s (TITLE-ABS-KEY search on self-serv* technology), by contrast, SST has been addressed second most commonly in the business literature (22,2%), with the Computer Sciences (22,4%) ranking first. With our selection criteria, the topics that thus far have been least addressed in combination are selfservice with augmentation and automation—even less so in the context of digital and service. For reference, there are almost 85.000 entries in Scopus on automation (with the same selection criteria), and 910.000 research and review articles across all disciplines—highlighting a gap in the literature looking at the connection of automation, augmentation, and self-service. The section below will detail the characteristics of the final sample and the procedure for data analysis. C. ANALYSIS Table II shows the final sample included for analysis. Selected papers provide a representative sample of the topic from the last two decades, ranging from 2005 to 2024. The sample consists of a combination of theoretical and empirical papers with 25 empirical papers, 4 systematic literature reviews, and 12 theoretical/conceptual papers. The sample covers various research disciplines including studies in management and organization, marketing and consumer behavior, information systems and technology, social sciences and public policy. Furthermore, there are sector-specific publications from healthcare, hospitality and tourism, public services, social services, retirement services, aviation, retail, finance, and professional services at large. While SST can refer to various types of technologies, most papers specify a type of technology with self-serving functions. The technologies studied range from information communication technologies (ICTs) and digital technologies at large, to self-service kiosks, voice assistants, robots, platforms, and AI in particular. The countries where empirical studies were set include a wide range from Northern European countries to Asia, South Africa, South and North America, resulting in a mix of developed and developing countries. As part of the analysis of papers, each paper was reviewed in-depth to explore how they approach automation, augmentation and/or SST individually and in connection. First, with respect to RQ1, this scoping process resulted in an overview of main topics around automation that the literature has commonly addressed: (1) the role of technologies for improving and changing work, job, and labor; (2) the relevance of human workers for specific areas; (3) the common resistance of workers to adopt to technologies due to a feeling of being replaced; (4) the changing and evolving relationships within and outside of organizations; and, finally, (5) the emergence of augmentation alongside automation. Secondly, with respect to RQ2, we reviewed and synthesized the literature further, focusing on emerging themes across automation, augmentation, and SST, building onto the first step. Through this structured process—collecting statements from the literature and forming groups—five themes emerged. As presented in Table III, some topics are commonly addressed in the literature across automation, augmentation, and SST. Those themes respectively move beyond those topics that are researched in each paper individually. As such, the resulting five themes highlight the interconnectedness of self-service, automation, and augmentation. These themes were further described and discussed. Furthermore, from the synthesis of the literature the conceptual model by Sampson [7], [20] was reviewed and adapted, discussing the role of SST in the presence of automation and augmentation [6] in services. To conclude, topics for further research were elaborated, based on the literature and emerging themes. TABLE II OVERVIEW OF SELECTED LITERATURE Author Journal Year Method a/Type Technologyb Sectorc Country [25] Technovation 2024 Qual VA N/A BR [7] J Serv Res 2021 Quan AI Professional US [6] Acad Manage Rev 2021 Conc/ The AI N/A N/A [2] J Serv Res 2018 Conc/ The AI N/A N/A [26] Soc Cult Geogr 2024 Qual SSK Aviation CN [27] Int J Soc Welf 2024 Quan RPA Social SE [28] Gov Inf Q 2019 Conc/ The Digital Public Services Public N/A [29] J Serv Manag 2018 Conc/ The Virtual assistant & service robot N/A N/A [30] Psychol Mark 2024 Qual Digital tech. H&T N/A [31] J Serv Res 2021 Conc/ The AI N/A N/A FIGURE 1. Scopus results for documents per year on automat* AND augment* This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2025.3620137 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ VOLUME XX, 2017 7 [32] J Res Interact Mark 2021 Quan AI Banking US (mostl y) [33] Soc Policy Admin 2015 Qual Web portal Retirement, NO [34] Soc Sci Med 2018 Qual Mobile phones Healthcare ZA [1] J Strateg Inf Syst 2015 Conc/ The Digitization, BDA N/A N/A [35] MIS Q Manag Inf Syst 2015 Conc/ The ICT N/A N/A [36] J Serv Manag 2019 Conc/ The ICT H&T N/A [37] Tour Manag 2025 Quan Robots H&T CN [38] J Bus Res 2020 Quan Platform Healthcare N/A [39] Int J Inf Manag 2021 SLR AI N/A N/A [40] J Mark 2005 Quan Interactiv e VR telephone system & Internetbased system Healthcare N/A [41] Manuf Serv Oper Manag 2007 Quan VR units, ATMs, and online banking Finance US [42] J Bus Res 2020 Quan Chatbot H&T Americ a [43] J Serv Res 2021 Quan Service robots H&T, Insurance, Aviation UK [44] Tour Manag 2021 Quan Service robots H&T Europe [45] Electron Commer Res Appl 2021 Mix Kiosk, Computer, Mobile Device Entertainment, Public Transport N/A [46] J Bus Ind Mark 2024 Conc/ The AI N/A N/A [47] J Hosp Tour Technol 2024 SLR Smart tech H&T N/A [48] World Econ 2024 Quant AI N/A Japan [49] J Travel Res 2024 Quant Service robots H&T World wide [8] IEEE Trans Eng Manag 2024 SLR Robo advisors Finance N/A [50] J Res Interact Mark 2024 Conc/ The AI N/A N/A [51] J Serv Theory Pract 2024 SLR AI Customer service N/A [52] Convergence 2024 Qual SSK Retail Germa ny [53] J Serv Mark 2024 Quant Robot, automatic machine H&T US [54] New Media Soc 2024 Qual SSK H&T China [55] Int J Contemp Hosp Manag 2024 Conc/ The Automated custo-mer assist-ant N/A N/A [56] Asia Pac J Public Adm 2024 Quant Disruptiv e audit tech Auditing ID [57] Gend Technol Dev 2024 Conc / The Digital tech HR N/A [58] Nord Welf Res 2024 Qual Digital SelfService Welfare NO [59] J Soc Welf 2023 Qual Digital self-mgmt system Public SE [60] Soc Policy Admin 2019 Qual Digital post selfservice Welfare DK aMethod/Type: Quant = quantitative methods; Qual = qualitative methods; Conc/The = conceptual/theoretical; SLR = systematic ltierature revieew; Mix = mixed methods. bTechnology: VA = voice assistant; ICT = information communication technology; VR = voice response; RPA = robotic process automation; SSK = self-service kiosk; tech = technology; BDA = big data analytics. c Sector: H&T = Hospitality & Tourism; HR = human ressources. III. RESULTS A. STATUS QUO OF THE LITERATURE (RQ1) The search for literature addressing automation, augmentation, and SST has resulted in a small number of publications from various disciplines but with a higher concentration in service research, information systems, business and management research. Surprisingly, literature addressing the three search terms at once appears to be rare. However, the review shows some topics commonly addressed that are: (1) how jobs are affected by the adoption of (digital) technologies, (2) the role of human touch and empathy in customer interactions, (3) the threat posed by technology adoption and resulting resistance by workers, (4) new and changing relationships within and across organizational boundaries as the service ecosystem adapts, (5) the emergence and growing attribution of augmentation. This section begins with a brief explanatory introduction to the three terms—automation, augmentation, and SST—and how they interlink. Subsequently, we discuss and present current engagement of scholars with the topics of automation, augmentation and self-service, and how SST affects service professionals and the service encounter. 1) FROM AUTOMATION TO AUGMENTATION AND SST At its core, automation refers to the introduction of technology, thereby shifting tasks traditionally done by humans to be machine operated. Today, the introduction of digital technology in organizations oftentimes leads to digital transformation, fundamentally changing how businesses operate and deliver value, requiring adapting entire ways of working [1]. Aside from operational changes, digitally savvy customers increasingly necessitate the development of new products and services. With vast information available online and freely accessible AI solutions, a range of services today This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2025.3620137 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ VOLUME XX, 2017 7 are thus gradually losing attractiveness, giving rise to digital self-serving applications without human touch. As a result, digital technology in services has significantly changed the delivery of services for both, employees and customers, shifting roles and responsibilities throughout [26], [52]. Although since the first industrial revolution, automation has mostly affected production industries and blue-collar workers, digital technology has exponentially affected knowledge-intensive and service-oriented sectors—not least since the Covid-19 pandemic. With more service encounters becoming automated and machine operated, however, scholars have noted the role of interpersonal elements and human touch [8]. AI technologies, in particular, further blur boundaries between human and machine, questioning the value added by each. As such, fostering a clearer division of labour, routine tasks may become machine-operated, while human workers may become essential touchpoints for customers. With the goal of enabling human-machine collaboration, rather than replacing human workers, machines become contributors or co-workers, whether in the digital or physical space. Alongside technological innovations from industry 1.0 to 4.0, we are now beginning to center our attention not on the technology but on the role of human workers. Although automation has advanced to impact physical and mental work equally, we have become progressively aware of augmentation where “humans collaborate closely with machines to perform a task” [6, p. 193]. Combined, automation and augmentation reflect approaches to technology adoption in organizations. In services in particular, however, technology often lies at the intersection of service user and service provider (or worker). The provision of timely, accessible and digital services therefore has resulted in a number of self-service solutions, automating various services. Yet, the replacement of human workers and full automation of such services often remains undesirable while augmentation currently misses clear strategies. In particular, many services continue to be built on trust, relationships (with workers), and receiving quality professional advice and care—aspects not always achievable with automated services—calling for augmentation. As such, recent developments and changes in service industries shed a new light on automation as we move towards increasingly smart (AI-driven) self-services, necessitating pathways for integrated human-machine service models and augmentation. 2) TECHNOLOGIES FOR JOB IMPROVEMENT With the first introduction of machines into the workplace, automation has been understood as a way to replace human labor. Then scholars have pointed to the significance of human beings in the workplace; a significant contribution in recent days is made by Sampson [7] who encourages to focus on job improvement rather than displacement. He further suggests that automation affects different tasks in different ways. Accordingly, first, tasks can be augmented by automation and remain with workers. Second, some tasks are deskilled by automation resulting in a transfer to lower cost workers. And third, tasks are moved to customers via SST, leading to a reduction or elimination of worker interaction. Despite various ways to automate, McLeay et al. [43] however calls for caution, finding how worker replacement can have ethical and societal implications for service providers [43]. Other means to automate, in parts aligning with Sampson, are presented, by Loebbecke & Picot [1] who distinguish five mechanisms where digitization and big data analytics affect labor. Through self-service, they argue, labor within physical processes can be substituted. High-level decision making on the other hand can be machine supported, substituting only some share of the work. In other mechanisms Loebbecke & Picot [1] share a pessimism for the relevance of human labor. Accordingly, knowledge work will be eliminated by data scientists and algorithms, and through new products and services, “physical forerunners” too will be replaced. As one last mechanism, more professional jobs will be split into smaller chunks of work overtaken by a growing mass of amateurs. Though some scholars believe in an increasing substitution of experts, they also suggest focusing on skills harder to automate, such as intellectual expertise, compassion, teamwork, ethical judgement, or cultural understanding. As described by Brady & Lin [26, p. 534], automation through self-service should be seen as transformation and “a synthesis and redistribution of roles” rather than an elimination of worker’s labor. In a similar way, Pentzold & Bischof call SST a “catalyst for restructuring labor” [61, p. 960]. Yet another suggestion for redirecting worker’s focus, as suggested by Breit & Salomon [33], may be towards increasingly complex work – a suggestion seemingly unmatched with Loebbecke & Picot [1]. Similarly, Altrock et al. [8] look at the interaction of robo-advisors, automated investment management tools, with human financial advisors, suggesting that dependent on technological ability, human advisors need to adapt their skills. As such, a less advanced robo-advisor requires human– machine collaboration (or augmentation), whereas a more advanced robo-advisor requires human advisors to reskill and redirect into other areas [8]. 3) VALUING THE EMPATHETIC SPACE While different strategies exist as to what happens to workers when automation is being introduced, scholars show a general tendency towards a focus on human-centered tasks. Youssofi et al. [30] empirically research digital technology adoption in the hotel industry, finding that when guests use digital technologies (i.e., SST), hotel professionals can enhance their competence as tasks become more efficient and cost‐effective. While guests experience autonomy, employees are then enabled to allocate time to more qualitative and human-centric tasks and make client interactions more meaningful. Rather than eliminating interactive tasks [7] or moving to technology related functions [1], the shift is then towards more meaningful This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2025.3620137 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ VOLUME XX, 2017 7 client interactions. Similarly, Lindgren et al. [28] note how “automation can lead to more interesting work”. As such, workers not only become bystanders but rather take on an engaging or a supportive role next to the technology. As described also by Brady & Lin [26] who explore the aviation industry, workers provide technical assistance and handle problems, thereby resembling the “empathetic space” to the technical interface. This may result partially in a reduction of interaction but more so in an expansion of their role into more meaningful interactions [7]. Despite the potential for automation technologies, SST in particular, to enhance communication and interaction among workers and consumers, they also risk affecting interpersonal relationships negatively. For instance, technologies may lead to a dehumanization of the service experience as users opt for technological solutions and workers are being replaced [36]. Furthermore, looking at the financial sector, Manser Payne et al. [32] add reduced opportunities for interpersonal communication and negative effects on relationships marketing as potential outcomes of automation. Furthermore, in public and welfare services automation and the use of selfservice applications can mean significant shortcomings for disadvantaged and vulnerable citizens [58], [60]. Here, automation may lead to dehumanization and depersonalization that generates barriers for users, restricting access to services they need [58], [59], [60]. 4) MITIGATING THE AUTOMATION THREAT In the long history of automation, the fear of replacement has been a significant concern, requiring efforts today to still think of its effect on individuals. Pan et al. [37] show what happens if outcomes of automation are neglected. With a focus on automation that disregards impact on employees, those may not only be de facto at risk to be replaced, as a possible outcome [7], but will even feel threatened, resulting in negative outcomes. Pan et al. [37] empirically explore robot adoption in the hotel industry. Yet, despite the potential benefits to guests and employees, they find that excessive organizational investments and high customer satisfaction can increase feelings of being threatened. This, as suggested by Pan et al. [37], calls for continuous recognition from the company and customers so that employees do not fear replacement – as organizations intent to retain workers. Across sectors the changes affecting workers will be quite diverse. As shown in the Indonesian public sector, workers certainly recognize the relevance for self-service solutions, although they believe their expertise gap to be very high [56]. In order to integrate solutions with workers thus requires understanding the differences and preparing workers to adapt. For instance, Persson et al. show how HR practices shift from street-level to screen-level, requiring masculine-coded practices (rational, instrumental) that were originally femalecoded practices (nurturing, supportive) [57]. And while repetitive tasks are the ones who should be first to be modeled [36], adoption of anthropomorphic AI and service robots become increasingly accepted by users, challenging the relevance of workers and possibly maintaining the threat if not managed [42], [43], [44], [47], [49], [51], [53]. 5) MANAGING NEW RELATIONSHIPS With the balancing of human workers and machines in service interactions, clarifying responsibilities is at the forefront of enabling automation. Next to ambivalent outcomes of automation for workers, the same goes for service users at the other end of the co-creating value chain [45], [46] [43]. Lindgren et al. [28] support but also criticize the use of automated systems in their study of public services. As they argue, algorithmic decision-making enforces an asymmetrical power relationship. Although in the hotel industry guests can feel increasing autonomy, “public services range from having a limited influence on a citizen's life to being of great importance to a citizen's economic situation and well-being” [28]. As also Germundsson & Stranz [27] caution, caseworkers in social services “might become less inclined to exercise their professional discretion,” once automation has been introduced. Thus, the automation of such services, and the different characteristics and unique qualities of human and technology-driven service provision should be carefully explored as the lines become blurred [28], [58], [60]. For instance, as noted above, some tasks such as compassion or ethical judgement [1] may be more human-centric than others and should thus remain with human workers [30]. However, contrasting an increase in asymmetric power and disadvantaging service users, Buhalis et al. [36] and Leite et al. [25] suggest that technologies also increase inclusiveness and accessibility to otherwise disadvantaged individuals such as the elderly and people with special needs who, for instance, may be physically restricted and might welcome digital applications. As traditional organizations can be slow to adopt, emerging organizations play an increasingly important role catering directly to consumers in need of accessible solutions. For instance, in the healthcare sector new market segments are developing, offering their solutions directly to consumers. As healthcare access can be difficult in some countries, emerging organizations provide self-care, preventive telemedicine, and disease prediction applications [38]. This allows patients to become more empowered and self-reliant without the need to go to a doctor. As Watkins et al. [34] highlight, in emerging countries certain gaps must be filled to augment healthcare. Thus, health workers and patients in South Africa develop their own digital communication tools, manage chronic diseases digitally, and access health information using the internet [34]. This shift, as Hermes et al. [38] argue, is moving an entire industry from linear two-sided markets to “complex interacting multi-sided markets mediated by platforms with super-modular/super-additive value creation” as SST becomes more than a “standalone feature” [41]. While patients, or individuals, had to go to a hospital or primary care provider, these roles are now being replaced and supplemented by This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2025.3620137 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ VOLUME XX, 2017 7 technology providers of digital health applications, as suggested by Loebbecke & Picot [1]. This application of technology is what Barrett et al. [35] describe as central for “the formation and functioning of service ecosystems” as resources are “combined and exchanged in new ways that create value for those actors engaged in the exchange”. Underpinning this shift in actors, Breit & Salomon [33] point to a gradual embeddedness of complex relations among citizens, frontline workers and digital infrastructures, moving away from the dyadic relationship. Similarly, Bolton et al. [29] describe an integration of the digital, physical and social realms, pushing organizations to acquire partnerships, network alliances, and finding mechanisms for aligning goals across actors. 6) ASSESSING AUGMENTATION As automation is still increasingly adopted, the shift in service interactions between human workers, customers, and technology intensifies—in particular with the introduction of AI. The majority of scholars are pointing to augment rather than replace [46], [47], [48], [55], as exemplified also by Pentzold & Bischof [61] in what they title “imperfect automation”. In 2018, Huang & Rust [2] define five stages of AI affecting jobs. While stages one to five suggest replacement of jobs with mechanical, analytical, intuitive and empathetic direction, the last stage depicts replacement or, more importantly, “complete integration with human workers”. Huang & Rust [2] name three factors for firms to consider automation with AI. Namely, the nature of the task, the nature of the service, and the strategic emphasis of the firm. A strategic focus is also brought up by Borges et al. [39] who suggest adjusting the AI–human equation “in alignment with business needs and digital strategies”. As a result, automation could be enabled that would generate competitive advantage for firms. Three years later Huang & Rust [31] present a continuation of their work, shifting increasingly towards the use of the word augmentation instead of replacement. Accordingly, different types of AI are existent with varying effects on human workers: mechanical AI, thinking AI, and feeling AI. As such, mechanical AI replaces human workers such as through SST, aligning on this aspect with Sampson [7] and Loebbecke & Picot [1]. Thinking tasks however “should” and feeling tasks “may” be performed by both, human and AI. The growing tendency towards augmentation has since evolved into somewhat of a necessity. As put by Raisch & Krakowski [6], “augmentation cannot be neatly separated from automation”. As we move “away from the question of what defines humans and distinguishes them from smart machines and towards querying how people and technology come together” [61], management perspectives have progressed. Put forward by Raisch & Krakowski’s [6] analysis, automation and augmentation are “interdependent across time and space”. Thus, through a coevolutionary process, humans and machines learn from one another, augmentation facilitates automation, and automation may enable augmentation. The determining factor for each application however is, as already defined by Huang & Rust [2], the nature of the task. In summation, automation, augmentation and SST are topics of vast interest by scholars from various fields and industries. Though some technologies replace human workers, scholars are increasingly pointing towards the redefining of roles or the improvement of labor. While some sectors are built significantly on human interaction such as public services or the hospitality and tourism industry, human workers remain present in many ways. As some tasks may be automated, human workers become the interface between technology and service user. As SST offers accessible service to users, service workers adapt, either moving towards technology-related tasks, empathetic tasks, or more complex cases. As AI continues to develop and digital service becomes increasingly available across markets, organizations that automate will have to strategize along the ways of automation and augmentation, keeping the human in the loop. Though this chapter has provided some topical overview, themes emerge across the topics of automation, augmentation, and SST that the following chapter will address. B. EMERGING THEMES FROM THE LITERATURE (RQ2) From an in-depth review and synthesis of the literature, five research themes evolve around the topics of automation, augmentation, and SST. While the previous section has focused on providing a general overview of automation, augmentation, and SST, the objective of answering the second RQ is to explore how the three are connected. In carefully reviewing the literature, we have found five emerging themes that provide an important interdisciplinary foundation for further research. Those are: (1) social (a)symmetries, (2) competing intelligences, (3) redistributed responsibilities and new skills, (4) merging spaces and new ecosystems, and (5) governance and ethics. Each theme described below highlights a fundamental aspect for enabling automation and augmentation through SST in service contexts. Table III provides an overview of how emerging themes were systematically derived from the literature with respective references. 1) SOCIAL (A)SYMMETRIES With the increasing adoption of SST, accessibility of services changes. However, SST adoption has two sides to this. On one hand, SST allows to democratize service and enable access to information and communication tools as can be seen in developing countries where, for instance, healthcare access is an issue [34]. SST allows to provide information and enhance services where those can otherwise not be easily accessed. On the other hand, SST does not represent a benefit to all, in particular, to disadvantaged groups. Few examples from developed countries show significant issues in providing selfservice options and automating services. For instance, welfare This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2025.3620137 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ VOLUME XX, 2017 7 services push the digital divide where disadvantaged or vulnerable citizens are becoming even more disadvantaged by these new services [58], [60]. Public and welfare services, as a result, become increasingly dehumanized, shifting responsibility from workers to individuals [59]. Thus, SST leads to mounting asymmetries as not all consumers have access or abilities to use digitalized services without the help of workers [27], [28]. Yet, often times, automation supersedes augmentation, resulting in job replacement or significant role changes. As a result, those unwilling or unable to adapt will be affected to leave their job. TABLE III OVERVIEW OF EMERGING THEMES Source Research Theme • Vulnerable citizens suffer with SST [58] • Consumer characteristics can affect the digital service experience [30] • SST makes disadvantaged citizens even more disadvantaged [60] • Automation creates asymmetric worker-client relationships [59] • Technologies increase inclusiveness and accessibility to otherwise disadvantaged individuals [25], [36] • Algorithmic decision-making enforces an asymmetrical power relationship [28] • SST increases availability of services, speed, and access to information at any time [33], [34] • Organizations must ensure that clients are not asymmetrically affected [27] Social (a)symmetries • Digitization and big data analytics may hit knowledge-based workers as hard and potentially faster than non-knowledge workers [1] • AI complements the work of teleworkers (RI) [48] • AI improves human work, intelligence automation (AI), by contrast, enhances human intelligence towards new things [46] • AI needs to provide personalized conversation and personal touch [51] • Different types of AI have varying effects on human workers: mechanical AI, thinking AI, and feeling AI [31] Competing intelligences • Digitization and big data analytics complement and replace labor differentially across industry sectors and work processes [1] • Abilities of human and technology need to be matched [8] • Consumers require role clarity, motivation, and ability for SST use [40] • With SST users become disciplined customers [54] • Customers become partial employees; but cocreation also entails co-destruction [45] • Users become empowered and self-reliant [38] • Workers are shifting tasks to help enable users’ digital literacy [60] • AI requires human oversight and accountability [62] • Services retain worker involvement for formal decisions [28] Redistributed responsibilities and new skills • Anthropomorphism increases acceptance of automation technologies [44], [53] • Imperfect automation requires new skills from employees and customers [52] • Automation requires a redistribution of roles [26] • Self-service reinforces responsibilisation of users [59] • SST moves tasks to customers, reducing or eliminating worker interaction and interpersonal communication [7], [32] • Knowledge work will shift to data scientists and algorithms [1] • Humans and machines learn from one another [6] • Employees move to human-centric tasks and make client interactions more meaningful [30] • Automation represents a masculine transformation away from female coded practices as nurturer and supporter [57] • Automation may result in a more engaging or a supportive role of workers [28] • Automation through SST is a synthesis and redistribution of roles [26] • SST frees up human resources for more complex tasks [33] • Citizens, frontline workers and digital infrastructures become gradually embedded [33] • Digital, physical and social realms merge [29] • Internet-based technologies facilitate new channels for communication [28] • Value creation evolves as resources are exchanged and redistributed among actors in new ways [35] • With designers and providers of technology a new group of actors is introduced [28] • Markets become complex multi-sided platforms with new ways of value creation [38] • Emerging organizations embrace significantly different value propositions than traditional organizations [38] • Interaction is gradually taken on by third parties [38] • New economies evolve as distributed networks, interacting with customers over platforms [36] • Service innovation in one dimension can result in changes in other dimensions within and across organizational boundaries [35] • Co-creation becomes difficult as solutions are disconnected [29] Merging spaces and new ecosystems • Worker replacement has ethical and societal consequences [43] • Confidentiality and ethics are relevant factors for effectiveness of digital service [30] • Self-service is not service, service requires personalization, expertise, experience, and empathy [55] • Self-service enforces depersonalization and dehumanization [36], [59] • Responsible service automation is needed to avoid bias [50] • Automation can be defined by the nature of the task, nature of the service, and strategic emphasis of the firm [2] • AI-human integration needs to happen in alignment with business needs and digital strategies [39] Governance and ethics This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2025.3620137 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ VOLUME XX, 2017 7 05, 2024, Emerald Publishing. doi: 10.1108/JHTT-06-20220159. [48] R. Baldwin and T. Okubo, “Are software automation and teleworker substitutes? Preliminary evidence from Japan,” World Economy, vol. 47, no. 4, pp. 1531–1556, Apr. 2024, doi: 10.1111/twec.13496. [49] M. Borghi and M. M. Mariani, “Asymmetrical Influences of Service Robots’ Perceived Performance on Overall Customer Satisfaction: An Empirical Investigation Leveraging Online Reviews,” J Travel Res, vol. 63, no. 5, pp. 1086–1111, May 2024, doi: 10.1177/00472875231190610. [50] W. H. Kunz and J. Wirtz, “Corporate digital responsibility (CDR) in the age of AI: implications for interactive marketing,” Journal of Research in Interactive Marketing, vol. 18, no. 1, pp. 31–37, Jan. 2024, doi: 10.1108/JRIM-06-2023-0176. [51] J. H. Blümel, M. Zaki, and T. Bohné, “Personal touch in digital customer service: a conceptual framework of relational personalization for conversational AI,” Journal of Service Theory and Practice, vol. 34, no. 1, pp. 33–65, Jan. 2024, doi: 10.1108/JSTP-03-2023-0098. [52] C. Pentzold and A. Bischof, “Achieving agency within imperfect automation: Working customers and self-service technologies,” Convergence: The International Journal of Research into New Media Technologies, pp. 1–19, 2023, doi: 10.1177/13548565231174588. [53] A. M. Barone, E. Stagno, and C. Donato, “Call it robot: anthropomorphic framing and failure of self-service technologies,” Journal of Services Marketing, vol. 38, no. 3, pp. 272–287, Feb. 2024, doi: 10.1108/JSM-05-2023-0169. [54] X. Xu and C. Greiffenhagen, “The disciplined customer: A video-based study of automated self-service hotels,” New Media Soc, vol. 26, no. 9, pp. 5013–5038, Sep. 2024, doi: 10.1177/14614448241251793. [55] C. Roberts and T. Maier, “The evolution of service toward automated customer assistance: there is a difference,” International Journal of Contemporary Hospitality Management, vol. 36, no. 6, pp. 1914–1925, Apr. 2024, doi: 10.1108/IJCHM-08-2022-1037. [56] M. R. Yahya, Jalaluddin, and G. Batara, “Current auditor expertise and future relevance of innovative audit technology: evidence from Indonesia public sector auditor,” Asia Pacific Journal of Public Administration, 2024, doi: 10.1080/23276665.2024.2312139. [57] M. Persson and A. Wallo, “Digital automation and working life of HR practitioners: a gender analysis of the implications for workforce and work practices,” Gend Technol Dev, 2024, doi: 10.1080/09718524.2024.2406039. [58] S. Hofmann, Ø. Sæbø, and H. H. Rydén, “Implications of Digitalised Welfare Services from a Vulnerable Citizens’ Perspective,” Nordic Welfare Research, vol. 9, no. 2, pp. 127– 141, Jun. 2024, doi: 10.18261/nwr.9.2.2. [59] M. Bengtsson, K. Jacobsson, and Y. Wallinder, “A journey into the new employment service landscape of responsibilisation: Towards de-personalisation of the caseworker–jobseeker relationship,” Int J Soc Welf, vol. 33, no. 1, pp. 137–150, Jan. 2024, doi: 10.1111/ijsw.12584. [60] J. Schou and A. S. Pors, “Digital by default? A qualitative study of exclusion in digitalised welfare,” Soc Policy Adm, vol. 53, no. 3, pp. 464–477, May 2019, doi: 10.1111/spol.12470. [61] C. Pentzold and A. Bischof, “Achieving agency within imperfect automation: Working customers and self-service technologies,” Convergence, vol. 30, no. 3, pp. 956–974, Jun. 2024, doi: 10.1177/13548565231174588. [62] W. H. Kunz, K. Heinonen, and J. G. A. M. Lemmink, “Future service technologies: is service research on track with business reality?,” Journal of Services Marketing, vol. 33, no. 4, pp. 479– 487, Sep. 2019, doi: 10.1108/JSM-01-2019-0039. [63] T. Fernandes and E. Oliveira, “Understanding consumers’ acceptance of automated technologies in service encounters: Drivers of digital voice assistants adoption,” J Bus Res, vol. 122, pp. 180–191, Jan. 2021, doi: 10.1016/j.jbusres.2020.08.058. [64] J. E. G. Bateson, “Self-Service Consumer: An Exploratory Study,” Journal of Retailing, vol. 61, no. 3, pp. 49–76, 1985. [65] A. Parasuraman and D. Grewal, “The impact of technology on the quality-value-loyalty chain: A research agenda,” Academy of Marketing Science. Journal; Winter, vol. 28, 2000. [66] A. Parasuraman, “Understanding and leveraging the role of customer service in external, interactive and internal marketing,” Frontiers in Services Conference, vol. 3, 1996. [67] M. J. Bitner, S. W. Brown, and M. L. Meuter, “Technology Infusion in Service Encounters,” J Acad Mark Sci, pp. 138–149, 2000. [68] W. J. Orlikowski and S. V Scott, “Working Paper Series ‘The Entangling of Technology and Work in Organizations,’” London, Feb. 2008. [Online]. Available: http://is.lse.ac.uk/ [69] W. J. Orlikowski, “Sociomaterial practices: Exploring technology at work,” Organization Studies, vol. 28, no. 9, pp. 1435–1448, Sep. 2007, doi: 10.1177/0170840607081138. SOPHIE ALTROCK received the M.Sc. degree in management of information systems and digital innovation from the London School of Economics and Political Science (LSE), London, U.K., in 2018. As a Marie-Curie researcher within the European Training Network for Industry Digital Transformation across Innovation Ecosystems (EINST4INE) funded by the European’s Horizon 2020 programme, she completed her Ph.D. degree in management at RMIT University, Melbourne, VIC, Australia, and at the University of Agder in Kristiansand, Norway. She previously worked as a Business Consultant at a multinational management and technology consulting firm, driving digital transformation projects. Her research focuses on the human side of digital transformation, exploring the impact of digital technologies on individuals in the workplace, human– machine collaboration, and value creation through automation and augmentation. ANNE-LAURE MENTION is a Professor and Director for the Global Business Innovation Enabling Impact Platform, RMIT, Melbourne, VIC, Australia. Her research interests sit at the intersection of technology and innovation management, business venturing and regeneration. Her research has been published in, e.g., Technovation, Technological Forecasting and Social Change, Journal of Product Innovation Management, R&D Management, and Small Business Economics. She is also a Co-founding Editor and Co-Editor in Chief of the Scopus-ranked open access Journal of Innovation Management, and has recently spearheaded a Henri Stewart Talk Series on Digital Transformation. Increasingly, the research projects she has designed and led use the SDGs as a lens toward intentional, responsible, and ethical societal impact. This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2025.3620137 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ 8 VOLUME XX, 2017 TOR HELGE AAS received the M.Sc. degree in information and communication technology management from the University of Agder, Kristiansand, Norway, in 2003, and the Ph.D. degree in strategy and management from the Norwegian School of Economics, Bergen, Norway, in 2010. He serves as a (Full) Professor with the University of Agder, and as a Professor II with Kristiania University College, Oslo, Norway. He is researching innovation management, and his research interests include topics such as the organizational effects of innovation, innovation processes and capabilities, collaboration for innovation, and management control of innovation. This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2025.3620137 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/