Knowledge Spillover: Between Serendipity and Strategic Planning - Lessons for Practice-oriented Interventions into Regional Innovation Systems

Previous CSCW research has paid increasing attention to regional innovation systems providing the context for CSCW research and design. In this paper we provide insights into attempts to intervene in such a regional innovation system via practice-oriented activities. Our findings describe experiences of co-creating knowledge around digitization of SMEs on a regional level, beyond individual company contexts. We reflect on the benefits and challenges such an approach affords and highlight the importance of interpersonal relationships and similarities between practices in different contexts to enable intended but also initially unexpected knowledge spillovers. Our findings highlight how a praxeological approach to CSCW can be influential in regional development programs, and open this field to CSCW research and design. We draw out lessons for similar interventions and outline fruitful avenues for future research.


INTRODUCTION
Over the past decades, interest in how research and design projects relate to the political, cultural, and economic aspects of the regions within which they are located has increased in CSCW and HCI.This includes research on how the concept of innovation relates to various geographical contexts, economic development and national identity [8] and how collaborative work plays into visions of manufacturing [28].The significance of rural areas has also become increasingly evident [22,[38][39][40]59].We add to this body of literature by looking at local learning processes and the difficulties encountered by Small and Medium-sized Enterprises (SMEs) in a rural yet industrialized part of Germany regarding digitalization.We explore our attempts to intervene in and influence this regional innovation system through practice-and action-oriented research and design initiatives.In doing so, we widen our originally rather firm-centered perspective, which is common in CSCW work, by addressing knowledge spillovers reaching beyond the boundaries of an individual organizational unit, firm or practice context.The nature of this intervention perspective complements existing studies of regional innovation systems in CSCW and HCI by providing valuable new insights regarding the importance of knowledge co-creation in such regionally oriented projects, the challenges of delimiting the effects of such work, and the necessity of building networks and relationships to facilitate its ongoing accomplishment.
The Federal Republic of Germany has a sizable number of rural areas whereby some of these regions exhibit at the same time a high degree of industrialization.This is where the bulk of the Small and Medium-sized Enterprises (SMEs) that make up the strength of Germany's industrial sector are located because they offer skilled personnel, the required space, and reasonably priced real estate.Known in the country as the German Mittelstand, these companies are often considered the 'backbone' of the German economy.Confronted with climate-related issues, the energy crisis and demographic change, there is an impetus to upgrade the historically energy-intensive production technologies in these regions and invest in the digital transformation of businesses, even if they still seem to be doing fairly well.
Rural areas present certain unique challenges, despite the fact that 'equality of quality of life' between urban and rural areas features in the German constitution (Art.72(2), GG).Many rural communities are experiencing population decline due to shrinking birth rates, migration to cities, and an inability to welcome and embrace newcomers or to provide appropriate access to the job market [84,86].A focus on traditional professional education that is less attractive to younger people, limited career opportunities inside the dominant SME structure, and cultural closure associated with high levels of 'bonding'-related social capital are seen as contributing to this situation [16].A number of measures have the potential to mitigate these concerns.One approach strongly advocated by the German government is to support SMEs' attempts to digitize their operations, procedures, and products through a network of so-called "Centers of Competency" (CC).These centers offer a range of services to businesses, including low-barrier educational events, training consisting of multiple in-depth seminars, and collaborative projects that use prototypes to facilitate the introduction and integration of digital applications in local companies.One such CC is the subject of this paper.
We anticipated a need for a context-sensitive approach with multiple stakeholders to properly understand the local innovation system [67].This matters because, while the programs and services provided by the center are often targeted at particular businesses, the regional innovation system in which they operate is made up of a wide range of actors.Such actors have tight, enduring ties, especially in rural areas.Different trade unions and their representatives, employer groups like chambers of commerce and industry, government organizations, and other public administrative institutions all play significant roles in this regional innovation system.This constellation of individuals and their interactions has to be considered in any initiative aiming to address and positively affect system outcomes.The center therefore adopted a practice-based and open co-designoriented strategy [30] that allowed for long-term collaboration [87] and involved key stakeholders in innovation activities [18].By tackling real-world challenges, this practice-based approach sought to meet societal demands and bring about social change [88].This paper presents findings, based on our experience of working in the center of competence, and examines whether such an approach is appropriate for regional interventions, rather than in an individual organizational unit or firm, and its potential advantages and limitations.
The methodological approach applied in this paper can be characterized as meta-research, or "research on research" [21,72] and incorporates the Center's own assessment efforts.This kind of research conception allows, or even partly requires, self-reflective observation and classification of actions by involved researchers which were in the observation period themselves part of the CC.It offers two contributions to CSCW.First, it reports on an effort to actively intervene in and influence a regional innovation ecosystem through practice-oriented CSCW activities, thereby broadening the existing debate on regional innovation in HCI.This offers a potentially valuable and fresh avenue of inquiry for HCI research, with implications extending beyond specific organizational contexts.Second, the focus on a practice-oriented approach enables us to discuss its advantages and drawbacks when concentrating on a region rather than a specific single practice-context, such as a department or production line.As the impacts of such interventions can "travel" through departments, businesses, and even sectors, rather than being restricted to their point of origin, we examine the importance of "spillovers" as an essential by-product of regional operations and how a practice-oriented CSCW might promote such effects.Our findings and the discussion provide preliminary lessons in this regard.

Innovation Ecosystems and the Role of Universities
Numerous studies have examined how universities contribute to regional innovation processes, notably in so-called high-tech regions (e.g.[2]).The development of Silicon Valley in California has attracted especially large amounts of attention because of the region's particular role in global technology development [2,76,85].This work has described how universities in the area evolved from largely independent research institutions into significant actors in a "triple-helix of industry, government, and university relations" [2,23], or highlighted the value of institutions like Stanford and Berkeley in producing highly qualified workers for surrounding tech businesses [76].Others have equally emphasized the pivotal role of Stanford University in transforming Silicon Valley from a rural to a high-tech region [2,85].The formation of spin-off enterprises that emerge from university research activities and generate cutting-edge technological products for local industries is another role that universities may play in regional growth.Although this approach has been seen as successful in some areas, it has produced more questionable results elsewhere, for instance in Northern Ireland [42].
The unique characteristics and difficulties of university-industry cooperation in research initiatives have been examined and emphasized in a number of publications.Organizations funding research initiatives increasingly demand the participation of other actors, particularly businesses, to increase the likelihood of transferable outcomes [21].Some argue that this has given rise to a new 'post-academic' paradigm for knowledge creation [90] or Mode 2 knowledge production [31,66].Dachtera et al. [21] describe a specific HCI research project that used a 'research by design' methodology.It involved three businesses and additional civil society groups and sought to provide technology solutions for certain use cases.They highlight a number of obstacles that arose, including organizational and epistemological differences between partner institutions.Similarly, Holdgaard et al. [44] point out the particular difficulties confronting Participatory Design (PD), with its bottom-up and open-ended methodology, when research funding systems mandate inflexible projects with pre-defined outcomes for research partners.These studies, however, focus on initiatives that only include a limited number of practice partners, making them quite distinct from the work reported here, which is attempting to have an impact over a whole area with large numbers of participating organizations and businesses.
Similarly, economics and related disciplines have long been interested in and focused on the dynamics of innovation processes in certain geographic locations, therefore a little detour into this debate is also necessary.Interdisciplinary network research has been a thriving area of study here [11,51,68].This literature notes a wide range of motivations for working with other network partners [49], including time savings [64], access to domestic and foreign markets [34,41], status and reputation building [33,70,71], access to knowledge [32], cost savings and risk reduction [34], and interorganizational learning [36].The relevance of knowledge transfer, co-creation of knowledge, and the collective nature of innovation processes have all been addressed, particularly by the neo-Schumpeterian approach to economics [37,50], which also established the "systemic innovation approach" [26,58,65].This viewpoint holds that innovations are the result of numerous knowledge-exchange and learning processes amongst diverse, strategic-oriented, bounded-rational agents inside a complex and highly dynamic system of innovation.Here, an innovation system is defined in a variety of ways, including a regional dimension (e.g.[19]), sectoral dimension (e.g.[60]), or technological dimension (e.g.[17].Since we focus in this paper on a rural-industrialized region, the regional innovation system perspective is most appropriate for the purpose of this paper.Our research usually supports the communal and collaborative nature of innovation processes and the significance of knowledge co-creation emphasized in this discourse, but it also highlights gaps, regarding for example the role of universities, which we describe below. One important concept in this discourse on how regional innovation systems work is the notion of spillover, which we also draw on in this work.The discussion around knowledge spillovers is by no means new.The concept relies on the notion that knowledge can travel from individuals to other individuals even across institutional boundaries.In economics and related disciplines, at least two types of knowledge spillovers are typically distinguished: 'Marshall-Arrow-Romer' (MAR) spillovers, and 'Jane Jacobs' (JJ) spillovers.The basic idea of the MAR concept goes back to the work by Alfred Marshall [61], Kenneth Arrow [6], and Paul Romer [75].It is argued that the geographical closeness of firms with a similar professional background (e.g.same branch of industry) facilitates the traveling of knowledge among these firms.The knowledge -e.g. on technologies, production processes, etc. -is carried and exchanged by individuals employed by the firms located within geographical agglomerations.Jane Jacobs [47], however, countered with a rather different argument.She argues that the geographical closeness of rather heterogeneous firms with different backgrounds and purposes is beneficial for knowledge to spillover and travel through space.Her argument is based on the notion that spillovers are most likely to occur in a regional agglomeration characterized by individuals with different perspectives, views and professional experiences, often associated with cities.Hence, JJ-spillovers emphasize that diversity of professional backgrounds provides a fertile ground for ideas and concept travel form one branch to another.The ongoing debate on the advantages of sectoral specialization of a region versus the advantages of a more diversified regional economy led to the development of the concept of related and unrelated variety [14,29].The 'related variety' concept specifies the 'Jacobs' argument by claiming the knowledge will spillover only when complementarities exist among sectors and firms experience a certain degree of competence overlaps [14].Particularly tacit knowledge is emphasized as locally embedded, context specific, and only transferable by the actors' engagement in mutual interactions, trust building and face-to-face learning processes [10].The role of technological artifacts, however, as carriers or vehicles of knowledge spillovers is not addressed explicitly in this research strand.
There is a large concentration of historic, frequently family-owned and -run industrial enterprises in the German Mittelstand where the CC under investigation is situated.While many of these businesses dominate their respective markets, they are frequently small.This makes it difficult for them to absorb lots of new students into their workforce, so as to augment their in-house knowledge of digital technologies.They also lack the funds to acquire technologies and technological knowledge through investment and the acquisition of innovative companies and startups, a common strategy in places like the Silicon Valley (e.g.[76]).We therefore focus here on how the co-construction of knowledge and novel socio-technical artifacts through cooperative encounters between enterprises and universities within a socio-informatic, practice-based HCI research paradigm can better serve this kind of geographical context.

Innovation Systems in CSCW and HCI Discourse
The operation of regional innovation networks in various nations and the function that HCI, CSCW and design projects play within them have attracted increasingly interest in the larger HCI community in recent years.For instance, several studies have emphasized the growing significance of 'innovation hubs' and maker spaces [28,55,56].These spaces enable technical and social innovation in ways that traditional research and development (R&D) laboratories cannot [56].These studies have spanned countries as diverse as Jamaica, Ghana, China [9], and India [45] [8, 20] and problematize claims that innovation models and methods originating in California's Silicon Valley, such as 'Hackathons', 'Design Thinking', or 'Start-Up Accelerators', are universally successful across cultures and contexts.In fact, by not taking into consideration local conditions and practices, these methods often fail to foster innovation, or even work against it by impeding local innovators' efforts rather than facilitating them [45,46].Such studies provide a significant addition to the CSCW community's understanding of the larger ecosystems in which its work is embedded, and serve as an essential context for the activities this paper describes.However, the majority of these studies make no attempt to influence these ecosystems and seek to simply describe and analyze them.This paper therefore adds to this body of work by reporting on what potentially happens when one does attempt to intervene in a regional innovation system.We reflect on the opportunity 'spillover' provides, and the importance of individuals occupying critical roles in the region and technological artifacts for creating 'spillover' opportunities which, if perceived, allow for CSCW work to transcend organizational and contextual boundaries, a highly relevant but as yet within CSCW unaddressed research setting.

The region
The region of Central Europe where our operations took place is dominated by small and mediumsized businesses, with a focus on manufacturing sectors such as automotive suppliers, machine tools, and electrical equipment.The local economy is facing regional-specific transformation challenges induced by developments such as climate change, demographic change, shifting global economic landscape, and a rapidly increasing demand for digital goods, processes, and services.This transformation has the potential for far-reaching social and economic repercussions for the region's inhabitants.
However, many local SMEs lack the conditions necessary for significant investments in digitalization, including funding and personnel.To address this situation, the Federal Ministry of Energy and Economy has funded 26 'Centers of Competency' (CC) to support SMEs in their digitalization efforts.The CCs provide education and practical, context-specific support to assist businesses in taking initial steps towards digitalization.These centers, often located in universities, are transferinstead of research-oriented and aim to co-construct digitization-related knowledge and assist in co-creating tangible outcomes for participating businesses.

The Centre of Competency (CC)
The Center of Competency (CC) under investigation in this paper is made up of a consortium of regional educational institutions, universities, and research groups.The major purpose of the CC is to close the 'digital gap' for SMEs in comparison to large corporations.The publicly financed program began in late 2017 and offers a range of activities for the innovative digitalization of local SMEs.The center focuses on human-computer interaction and emphasizes the importance of employees in digitization efforts.The center's primary areas of concentration include the adoption and use of digital technologies in businesses.It is however not the technologies themselves that are of importance, but rather how they are used and appropriated, together with the resulting adjustments to organizational structures and the larger innovation system.The people running the center believe that employees should play a central role in digitization measures, as their knowledge and practices are essential to both the success of the company as well as the design and appropriation of technological artifacts.Particularly, workers' innovation, current experience, knowledge, and competencies -many of whom in this region have been employed by the same company for many years -are thought to be essential to the success of the businesses.Digital technologies should therefore support this qualified human work rather than standardize, automate or ultimately replace it [57,88].Additionally, the center strives to create a type of digitalization that is mostly employee-centered and embraces the spirit of 'industrial relations' [35] in line with the CSCW community's broader interest in (work) practices [53].

Epistemological and Methodological Foundations of the Center
All efforts are founded on a commitment to the central role of practices in CSCW and HCI research [12,52,78].Following such a practice-oriented paradigm, technological artifacts are appropriated by users to their specific context [69].Thus, technologies get enmeshed in human activity and behaviors, which is a dynamic, complex, and contextual process [1].'Use' is an active and creative process, as those who apply and appropriate artifacts give them meaning in their specific context of application [69,74,79,80,82,83].To address the disparity between designer's understanding and user's real use-context, research is divided into 'design case studies, ', which include an iterative combination of empirical investigation, participatory co-design-intervention and investigating and supporting processes of appropriation of the created artifact to the specific context [81,87,89].This iterative and participatory process aims to overcome the 'asymmetry of knowledge' [24,73] that exists between the various actors, such as academic researchers and practitioners, and to create shared knowledge about the practice context at hand and the design possibilities [25].
The Center's operations attempt to intervene in such behaviors, primarily through the deployment of socio-technical artifacts, with a commitment to seeing technology as socio-technical artifacts given meaning through usage, and the demand that individuals of the settings in which we operate participate in the digitization process.Knowledge sharing is also fundamental to the center's operation, since it is entrusted with increasing SME's ability to address and successfully execute digitalization initiatives.To ground our approach to knowledge creation we draw on socio-cultural theories of learning (see, for example, [54] and following the lead of [24].Learning at the CC is thus viewed as a process of co-creating knowledge among many actors, with no predefined and rigid roles of expert and student.In this approach, knowing is contextualized, mediated by artifacts, and diffused across the social environment, with no single individual possessing all relevant information.The center's research focuses on incorporating diverse perspectives and knowledge into sociotechnical interventions through participatory approaches, with the goal of creating value for regional rural SMEs.Studies are done through action research [43], where knowledge is generated through intentional intervention into practices.This approach is suitable for technology transfer projects where the primary goal is to create value for practice partners rather than scholarly knowledge.

Activities of the Center
The Center of Competency (CC) developed and offered a variety of initiatives for and with SMEs as well as other actors.Their program is divided into three categories: 1) informational events, 2) workshops, and 3) digitalization projects.
Informational events are low-threshold events in which participants learn about the issues that firms may face as a result of digitalization and how they may respond.These activities frequently provided participants with their first contact with the Center's program and informed them for the first time about the many assistance measures available.
Workshops are more intensive events, often bundled into thematically cohesive workshop series, one of which is named digiXpert 1 .These workshops provide participants with experience specifically relevant to address challenges directly inside their own firms.Real-world examples, such as concepts or difficulties connected to digitization that participants encounter in their own organizations, play critical roles and provide the examples around which collaborative workshop activities, such as co-design activities, are undertaken.As a result, these seminars primarily adhere to the commitment to a practice-oriented approach outlined above.
Digitization projects are collaborative projects that frequently expand on the difficulties and ideas presented at the workshops, which are typically executed by personnel of participating firms and university employees.However, it also happens sporadically that digitization projects are requested and initiated 'out of the blue', i.e. without prior contact by an SME with formats of the CC.As we will elaborate on in more detail later, these -at first glance spontaneous and unexpected 'digitization project' requests -are by no means purely coincidental.Although these design projects often last only a few months, they employ a participatory and practice-based approach.This implies that members of specific practice settings, such as a department within a corporation or a specific manufacturing line, take center stage, and any interventions begin with ideas, problems, or requests stated by them.The fundamental purpose of these design cases is to investigate and show IT-opportunities to assist qualified work in a focused manner and co-create knowledge and expertise inside the organization to carry out their own digitization initiatives.

METHODOLOGY: META-RESEARCH
To describe, capture and analyze our own activities, we conducted research on the above introduced project as a form of meta-research [21], often as part of the formal evaluation activities we were carrying out.As previously stated, the CC under inquiry began operations in late 2017.Following an initial setup phase, the activities detailed in the preceding section began immediately after the project's launch, in 2017.By 2021, four rounds of the digiXpert workshop series had been completed, 28 design or implementation projects had been completed, and members of the Center had organized or participated in 421 events.The research process can best be described as a form of action research [43], as it is rooted in the intervention of the CC with the aim of creating social, technological and economic change in the organizations the competency center works with and the region as a whole, as to the goals of the CC described above.To understand the work of the center better and reflect on their own work, the members have thus studied their own activities and practices.To this end, since the commencement of the project, the researchers involved have kept research diaries, taking short notes during certain events, steps, and meetings that were eventually expanded to full field notes [5].These include, for example, notes gathered during the digiXpert workshops or transcripts of interviews with design project participants as part of the collaborative study into specific practices in order to collaboratively create design solutions.Second, as part of the Center's internal review, qualitative open interviews were held with members of organizations that had participated in the Center's activities, as well as members of the Center.Such interviews are done on a regular basis following the completion of the corresponding tasks.Apart from identifying potential areas for development in the Center's program, these interviews also served to obtain a deeper, more complete knowledge of the Center's activities, their perception by partners and participants, and potential inter-dependencies and synergies.This assessment is continuing.Nonetheless, it should be highlighted that data gathering was somewhat opportunistic, since the project's sponsoring organization specifically opposes any involvement in 'research' by center personnel, and research efforts must thus fall within the center's and its customers' aims.As a result of this meta-research practice, the authors of this paper are also involved in the efforts of the center in various roles, including parts of the activities and experiences detailed in the following section.Some authors are not involved in any of the activities this article describes, but are nevertheless part of the center.Some of the efforts described below involve thus none of the authors, but were documented during the meta-research the authors carried out.Other activities described below do involve some of the authors of this paper (but not all).We have tried to make such involvement clear in the narrative presented in section 5.
The material gathered from the field notes and interviews was integrated for the sake of this study and the reflection of the Center's work that it includes.This data was then thematically analyzed collaboratively [15].Authors compared and discussed their findings, and when required, sought clarification from members of participating organizations.This procedure resulted in the narrative(s) offered in the following section.To maintain anonymity, all names of actors, organizations, locations, or events that might be potentially revealing have been changed.Any names that appear are pseudonyms.

FINDINGS
Our findings are separated into two chapters.The first chapter tells the story of two cases in which regional actors participate in the digiXpert series and how this participation ties into the distribution of effects in the region.The two cases show how the distribution of effects of the center's work is aided by the personal/professional networks of employees of the center and their further contacts.The second part describes the work employees of the center undertake to create and maintain these connections.

Distributed effects
In this section, we detail two cases, which are similar in that participation in an instance of the digiXpert series is a central experience for the protagonists.The first one describes how a project initiated by members of the center with an industrial company ends up also benefiting companies in the care sector of the region, aided by the networks of a mediating agent.The second case describes how a person becomes such a mediating agent for the work of the center after participation in the digiXpert series.At the center of both cases is a check list application.Together, the cases illustrate how the effects of interventions become distributed through the region.
5.1.1The development and deployment of a checklist app.Georg2 , an employee of a manufacturing company participated in several of the Center's programs which finally enabled him to drive digitization within the company.The company he works for is a larger industrial company with several subdivisions, focusing on the production of parts for the automotive sector, environmental technology, as well as tools.The local production site with which we cooperated employs about a thousand people.It is thus not an SME, but a larger company with many subcontractors in the region.Thus, projects at this place have a potential 'lighthouse effect', with smaller companies paying close attention to such developments.Although the funding program that finances the CC is originally aimed at supporting exclusively SMEs, the funding authorities are of course aware of potential 'signaling effects', hence, accept activities with large and visible companies.
Georg first heard about the Center at an informational event the Center participated in.He had a background in computer science.Part of his responsibility at the company included supporting the internal lean management process.The event he participated in took place recurringly in the evening under the same title and with the same format, but each iteration focused on a separate topic, with experts as well as company members presenting insights and experiences with specific technologies.Georg remembered that "at these events, new developments, technologies, services are presented and discussed, and at this specific edition this included the program of the CC.That's how I heard about it and I got in contact with the coordinator of the Center".
He decided to seek further engagement with the Center and signed up for an iteration of the digiXpert workshop series.The series included seven individual workshops and covered several topics in a hands-on fashion, ranging from process modelling, change management to the capture and analysis of machine data.This hands-on and applied nature of the workshops was perceived as especially valuable by the participants: "At the workshop I could see how with rather simple prototyping means you could connect sensors and collect some data.This I really liked a lot", Georg told us.
One of the main outcomes, however, was not the direct application of the lessons at the company, but an idea for further cooperation between university and company: "The result of the prototyping workshop was, that we started to discuss how we could realize a small project together in our company.Then this idea with an app for machine preparation came up.Something, that could be done quickly and work well. . ." At the company machine set-ups have to be regularly changed for different production steps.These preparation processes are constantly improved and hence changed frequently.These changes are difficult to document in an up-to-date manner with analogue means, and employees struggle to get accustomed with new processes and not resort to old and well-known patterns.Furthermore, process technicians struggle to study with the necessary granularity how processes are actually carried out and to identify where in the process problems occur.A digital application was thought to alleviate these challenges.
Subsequently, members of the Center and Georg began working on a 'smart watch' application, that would guide employees step by step through setting up machines, while also allowing to measure the time each step took in order to identify where workers experienced difficulties, which steps took significantly longer and where the optimize different the set-up processes.This was a set-up as a co-design project, where the participating company members provided domain expertise about machine set-up and the company environment, and the employees of the center provided technical expertise.Several visits were organized to the company site, to observe machine set-up processes together.The co-design team consisted predominantly of Georg and several members of the center.When it was decided that a smart watch would be most suitable, as workers hands would remain relatively free during the set-up process, another member of the CC was brought in with specific expertise in smart watches.However, due to his knowledge in computer science, Georg also took on some development work himself."Whenever I had a little time, I played around with it.[...] I had to re-write a lot of my code when I got stuck.But that was also my ambition.It was only difficult in the sense that I had never done it before.It was the first time I helped programming an app, why would I have done it before?(laughs) But I didn't find it difficult, but as a chance to learn something." The application prototype underwent several iterations.A first prototype was presented to workers, the head of production (Jan) as well as the work council (Moritz) 3 .As the application was programmed without any collection of personal data, Moritz the work council approved.Workers demanded that the font should be bigger to improve readability.After these changes were implemented, the application was considered ready for testing on-site.In that moment, however, unforeseen challenges emerged.The application was running on a server at the university, a setup which was deemed against the company's security protocols, which demands that applications are run from servers within the company's premises.This provided a hurdle insurmountable in the short time available, and put a rather abrupt end to the project.
Nevertheless, the project was beneficial to Georg and his company, they reported to us.The collaborative development process, the new coding experience the employee gained and the illustrative potential of the smart watch application led to significant changes within the company.Despite the inability to test the application in situ, the process enabled the company to gain significant learnings."It was nice to see that it does not always have to be a fantastic 1000 € solution.To see if something provides and added value, a cheap 80% solution is sufficient.It [the project] simply gave us the impetus to think differently.[...] Usually we only implement applications when we are very sure that everything will work perfectly", Georg shared with us.
This new way of thinking had also very direct and structural consequences for the employee and the company.Georg was given the new responsibility to drive digitization efforts within the company: "The plant manager then [after the project] decided that I could focus on such (digitization) projects." The experience he gained in the collaboration with the Center prepared him well for this new role: "Because (...) now I simply have more knowledge and can then suggest to our plant manager 'Let's make a quick prototype for a second, it'll go pretty fast.' [...] I also don't try to push the topic of digitization into our company.Instead, I try to ensure that we simply have good tools for the problems we face".Not only his own attitude and confidence towards digitization has changed, but that of the entire company: "Before, we were really only ever driven by our customers: For example, the customer said you need a connection to our ERP system so that we can send the orders directly through to you.Then we just did it, but we were not the driving force".
The project then continued to create effects in other contexts of the region.On the one hand, the checklist-like system the application implemented provided the foundation for another implementation project in a different industrial company, where it served to create a quality management application, which also employed a checklist, similar to that used to control machine set-up.The application was appropriated collaboratively by the company and members of the CC to facilitate the assessment of the quality of large welded machine parts.This was up to then done via a paper questionnaire.Through the project the smart watch application was transformed into an Android-based tablet application, which the quality manager could fill out the checklist digitally and include, for example, pictures and videos to improve documentation (see our other paper for a more detailed description of this case [reference omitted for blind review]).
Furthermore, these checklist-like applications later found application in other companies, as well as far away from the industrial context of their origin, in the care sector.We will briefly describe these developments below.

5.1.2
The check list app is distributed through the region.To enable the transition of the application from industry to care, the impetus and networks of a mediating person from the regional administration were instrumental.For the sake of this paper, we call this person Ulrike.
Ulrike is a member of the regional development office of the county the Center is located in.Her role includes both economic development in the region as well as coordinating support for the health care sector.Next to larger institutions such as hospitals and government organizations or NGOs, this sector also includes various smaller private care institutions which are SMEs.In 2018 Ulrike organized an event on the "future of care" in the region.The main group of invitees were professionals from the care sector, but the guestlist also included technology providers and Bernd, a member of the center for competence.Invitations were sent out via email to Ulrikes own professional network and in some cases additional phone calls were made.Bernd was invited to present the format of the digiXpert series specifically to the care sector, and to join a workshop which was part of the event.
Part of the workshop was a discussion on digital technologies available to support documentation duties within care.In this discussion it became clear that documentation included a lot of routine activities, such as the manual documentation of patients' vital signs several times a day or noting the temperatures of fridges storing medicine.Bernd, participating in this workshop, suggested that the load of some of these routine activities could be supported with the previously developed checklist applications, which could easily be adapted to the care context.The suggestion was met with enthusiasm by the participating care professionals and a working group formed to explore this opportunity further and develop a prototype.Ulrike, our contact point in the regional administration who had organized the event, was chosen to head this working group.
Once the group began to meet regularly, they quickly decided that a checklist-like application would help to eliminate pen and paper and the later digitization of manually collected data.A smart watch was deemed to be the most suitable device for such an app, as it would keep the hands of care professionals free, just like in the industrial company, where a smart watch was used during machine set-up procedures.The group then began to visit care facilities, where it quickly emerged that a smart watch was not at all feasible, as it violated general hygiene regulations, and smart phone would be better suited.A prototype was developed and discussed in another workshop organized by the head of the working group.
However, further visits to care facilities revealed that many of the devices from which data needed to be recorded, such as fridges and pulse oximeters, had interfaces that would allow for automatic collection and transmission of data to a central data bank.It would thus not be necessary to fill a digital checklist.Developing a system that would either enable automatic recording of data from either pulse oximeters or from fridges and ambient sensors in patients rooms to assess general hygiene conditions became the new objective of the group.
The Covid-19 pandemic put an extended break to these efforts.However, one of the participants, Marlies, continued to lobby for the project even during the pandemic.Whenever possible Marlies presented the work of the group to other care professionals.Through these presentations she initiated a cooperation with another care facility she was in contact with, who had previously not been involved in the group.This new care facility provided the opportunity to put the conceptual prototype into practice.With the support of the CC several devices, recording environmental and health data that care staff needed to record everyday, were connected to a locally hosted platform via their existing interfaces.The data was then sent automatically to a databank and displayed on a dashboard, and staff no longer needed to record manually, walking around the care facility.Although the implementation of this project included the involvement of the CC, it was fully initiated by Heidi, who contacted the care facility, convinced them of the potential benefits and only at the last step invited Bernd and his colleagues to assist with the implementation.
Ulrike, who had initiated the project, also did not stop her efforts.She regularly, whenever pandemic measures allowed and with renewed effort afterwards, presented the process and especially the digiXpert series to members of stakeholder organizations such as care facilities, chambers of commerce and industry as well as politicians.She also targeted other regions, to broaden the range of influence and convince others of the benefits such workshops would have for the care sector.This activity highlights the value the cooperation between the competency center and mediating organizations such as this administrational office can bring for the involved partners as well as for participating SMEs.A statement by Ulrike underlines this potential: "The two of us, Center and my office, playing together is crucial for the region.[...] It helps us and you, and drives connections in the region.It has become a network and it is vital to carry it forward, as both sides can endlessly benefit from it." During the pandemic, another application context for a digitally-represented checklist emerged.The head of the center of competency, Markus, was contacted by the Tom, the head of a small company that offered the assessment of work place safety for various kind of companies.His specialty were the assessment of fire safety in factories as well as high storage shelves.Companies are legally obliged to conduct safety checks, follow safety protocols and receive external documentation about their fulfilment of these obligations, which is exactly what this person offered.Tom's role was to visit companies in need of safety audits and assess safety using paper-based checklists.The new request from the SME was forwarded via internal communication channels within the CC to those staff members who had already dealt with the checklist app in the past.The CC was able to build on the existing system and together with this person appropriate it in a way that it was suitable for his specific work context, using a tablet when visiting companies.

The work for the network
As the narrative above illustrates, making effects take hold in the region, making artefacts and knowledge travel across contexts, relies on networks, personal contacts, exchange and collaboration between people.To create this network, these connections, requires work.In the following we detail some of the activities members of the center undertake to create and maintain these connections: the work they do for the network.
These activities, for the sake of this paper, can be split into two larger categories: one is networking, the other one is public relations work.

Professional networking.
Members of the CC revealed that they actively, purposefully and continually engage in activities to connect with other actors in the region and to maintain these connections.These include visiting events, in-person meetings as well as phone calls or online meetings.As Fritz shared with us: "When I was new in the center I used to visit events just to meet people.For example, I would drive to [CITY X, a larger city about a 90min drive away] to participate in an event in the evening to build my network.[...]".These events are organized by other competency centers, chambers of commerce, the office for economic development or local union offices.Fritz would go with the sole purpose of connecting to people as a member of the audience.Sometimes, they revealed, they met people there that they already knew, but often times it would be new people that they would simply approach "in the coffee break of the event".
Nowadays the situation is different.Fritz and other members of the CC still participate in events, but these visits are more directed."Nowadays I know people", Fritz told us."I still go to events, but I go there because I am part of the organizing team or because I know that others will attend who I want to meet.Often these are people who I already know, to see them again and maintain contact, maintain my network." Maintenance has thus increasingly become a more important focus, albeit not the only one."I also still want to expand my network.I get to meet people whose name I already know because I have heard of them elsewhere, but I have never met them, there is no personal contact, which then happens at these events".
Events are not the only manner to maintain these networks.Maintenance activities also include meetings.Often, these meetings serve no other purpose other than to stay in contact, without any concrete objective or outcome."I just want to hear what people are working on, what topics are on their mind currently, just to stay in contact".
Other times, these meetings have a concrete outcome, such as the planning of joint activities.An example includes a repeated exchange with a local business consultant around the topic of 'new work', to plan and organize joint events.Sometimes, these networks and associated activities take on more formalized shapes.For example, a network exists that calls itself the working group 'Knowledge Management' within SMEs.It exists since 2019 and consists of a core group of five people, including two members of the center (Fritz and another member), a member of a local vocational institution (Sabine) and two members of SMEs.Their objective is to organize workshops and informational events at least once per quarter.For these events the core group asks others, external speakers or hosts that are able to share specific experiences or expertise.Topics include, for example, the role of AR/VR in supporting knowledge management, knowledge management in relation to ERP systems or processes to pass on knowledge within an SME."The work [to organize the events] is thereby distributed across more shoulders.It is less work but great workshops."The invitation to the workshop is then distributed via the professional or personal networks of the members of the working group.For example, Sabine sends it to its large network of member companies.
"Using other organizations for their networks is extremely important, in general, as the university, especially in the beginning, did not have a strong network" we learned from Fritz.Other organizations, such as unions or public economic support units have long standing and large regional networks, that become accessible through cooperation, such as exemplified in the working group.
Not only organizations are crucial, but regionally well-connected individuals have repeatedly been proven to be instrumental in establishing new connections and relations for the center, which enable new projects or other activities.Such individuals include Ulrike, our contact from the regional administration mentioned above, who enabled the checklist application to move from the industrial sector to the care sector, but also the Sabine of the educational institution for SMEs, a member of the working group on knowledge management.Her cooperative activities are not limited to this working group, but she has become somewhat of an advocate for the competency center.Before her participation in the working group, she participated in an iteration of the digiXpert series.
Sabine's employer offers vocational training for apprentices of companies in the region.Vocational education is an important training path for regional companies.The center offers seminars as well as training facilities with a large park of specific industrial machines.Beyond these activities for apprentices the center is further also responsible for offering general educational programs for regional companies.With regards to this second responsibility of her organization, Sabine saw great potential for synergies with the competency center, especially with the digiXpert series.At various occasions, usually events organized by other organizations such chambers of commerce or trade, she held presentations about her experiences as a participant within the workshop series and answered questions from the audience, taking on the role of an ambassador for the Center's digiXpert workshops.These events served to generate interest amongst a broader audience, leading attendees to register for an iteration of the workshop series.
Over time, this collaboration increased, leading to jointly organized events between her and the members of the CC, apart from the work already taking place in the knowledge management working group.These events, usually held at the premises of the educational service provider she works at, aimed at furthermore bridging the university and local industry by also including companies from the region as speakers, which are the clients of the organization Sabine represents.Together they offered a series of workshops.They targeted specifically trainers in the companies as well as members of HR departments and covered various topics related to digitization to provide a broad overview of the topic, including, for example, workshops on IT security, change management or the use of sensors to measure machine performance.
The organization of the digiXpert series further illustrates the importance of these network of networks.Here, we learned that the first series was advocated afterwards by the federal ministry of work, health and social matters within their network.Subsequently the center was approached by other organizations such as the offices for economic development from (sub-)regions to organize more series, and since then the series has been iterated many times.Furthermore, other centers of competency and other organizations started to copy and appropriate the format for their own region, sometimes with a bit of initial guidance by members of the CC at the core of this paper.
Overall, it becomes apparent that 'networking' is an amalgamut of activities with the sole goal of increasing and maintaining specific professional contacts, but also that these contacts are crucial for the work of the center, as they enable the center to reach larger number of possible participants, especially SMEs.
Another set of activities has a similar purpose, but takes the shape of public relations (PR) work.We will now briefly turn to this work.

Public relations. Public relation (PR
) is a crucial component of the CC.Its primary target group are SMEs.The aim is to communicate the program of the center, success stories of projects with other companies as well as knowledge and other outcomes generated through workshops, events or design projects.As a secondary target group it is directed at press organs such as newspapers or individual journalists, as vehicles to increase its reach.This work also consists of specific, directed activities.
For example, the center maintains an internal working group on social media, consisting of one member of each consortium partner.The group meets bi-weekly for one hour to share recent activities from each partner, discuss which one is suitable for publication and develop an editorial schedule.This schedule is then executed by the dedicated PR person of the center, Heike.For Heike, the meetings have an additional secondary purpose.They serve as a key moment to learn which topics other consortium members are working on, which topics receive attention by SMEs and which are not very popular.This knowledge then serves other PR work, such as blogposts, podcasts or brochure development and distribution.
Other sources of input for PR work include bi-weekly consortium meetings, but as well as individual meetings.These are mostly phone calls with other members of the center, sometimes also unscheduled when a specific question or issue needs to be resolved quickly."Usually, there is a concrete issue, but then other things come up, and one of us says "now that we are talking anyway, I also wanted to ask you about XYZ" and then we talk about what they are working on", Heike explained.This kind of contact happens regularly."It is a crucial part of my work, because only this way I gain access to the knowledge I need", she told us.Meetings also take place in person.Heike told us that she and colleagues also at times visit partners across the region to meet, especially to look at something, for example a specific hardware installation.While other sources, such as (social) media are also used, "personal contacts are always more important than social media for example." While the key target group of this work are SMEs, it is aided at times by mediators such as journalists.She also told us that she maintains a network of contacts to such mediators, and meets them for a coffee.Often these meetings are about a concrete story, but not always."These meetings are always super helpful, insightful, but it does not happen so often, I do not have the time, and journalists are also busy." Another crucial partner is the PR office of the university.Press releases are often published through them, as they have and maintain a larger network than the center by itself, which enables the CC to reach a much larger section of the target group.
Just like the activities detailed in the previous section above, PR work is also a form of networking work that serves an important role for the center.Information about the center needs to reach regional SMEs and personal contacts are not sufficient to reach the entire region.PR work assist the center to inform companies about their existence and the support they offer.This way, new contacts and projects are generated.For example, after a press release about ongoing projects that explore the circular economy for industrial companies three additional regional companies contacted the responsible center member with a wish to be involved.

DISCUSSION & LESSONS LEARNED
Previous studies from within the HCI community have investigated and reported on innovation systems as context for HCI work, such as work by Avle, Lindtner or Freeman (see e.g.[8,9,27,28,55,56]).Such studies have been crucial work for the wider HCI community as they initiated the discourse and investigation of innovation systems and regions as context for design research work.Our work intends to develop this discourse further by providing a report on an attempted intervention in such a regional innovation system.Our findings exhibit different ways in which the efforts of the CC initiate change in the region, affect SMEs and involve a variety of actors, including company employees and middle managers, university researchers, representatives of chambers of commerce as well as members of public administration.
The experiences described above hold several preliminary lessons for CSCW research targeting regional innovation systems.Firstly, it becomes clear that the effects of our interventions possibly extend beyond the borders of collaborative research and design projects with individual companies, which are arguably a common form of practice-oriented CSCW work.This is exemplified by the roles Ulrike and Marlies took on to carry the results of previous work within an industrial company to the care sector in general and then via Marlies' networking to specific care facilities where such previous technology and knowledge it represents is appropriated to this new and different context, to which the CC initially had less access.This initially presents somewhat of a methodological challenge for a practice-oriented approach, as participants and their relations extend significantly beyond the borders of what is usually understood as a specific practice context.The question emerges how CSCW research and design projects can take the complexity of contexts into account, and account for their own role in them.We believe our experiences provide some insights towards answering this question, highlighting the central role of practices also in region-oriented work.To be clear: we do not mean to present collaborative workshops or design projects as a new approach.Quite to the contrary, we mean to show how such established praxeological approach to CSCW research and design work can play a role in regional efforts, that go beyond a single practice context such as an individual organization or firm.This depends on several factors, including similarity of practices between contexts and the role and networks of members of the CC but even more importantly the networks of other actors, local champions such as Ulrike, Marlies and Heike.These networks that extend across the region are vital for the spillover we described in the previous chapter to occur.Secondly, and following from the previous lesson, a regional focus seems to require increased flexibility from university and research staff in what are considered appropriate activities for university members, in order to deal with the necessary relations and reputation, which can pose further significant challenges.Our findings also enable reflection about these challenges and how they open up potentially fruitful avenues for future CSCW work.Together, we believe, these lessons open up regional development as a potential field for praxeological CSCW research and design.

Effects travel through the ecosystem
The findings of our inquiry, as described above, show how the effects of design interventions, such as participatory workshops, are not constrained or limited to a specific practice context or ensemble of actors.This is arguably true in a general sense when intervening through research and design, even if our target group or circle of participants is intentionally constrained.In her work on Locating Accountability, Lucy Suchman, for example, makes a similar point [83].In our instance, however, de-limiting the impacts is precisely what is sought, in order to address not just individual practice settings such as a department or company, but also to produce effects that extend throughout the region.The region we are focussing on is characterized by a historically rooted, multi-dimensional, and rather complex network structure.Actors are interconnected in multiple ways.Engaging actively in such a network environment comes along with the emergence of unanticipated repercussions and effects of interventions.Hence, we argue that spillovers "travel" through these network channels, sometimes intended, but often also unintended.
What actually travels in the example given in the first subsection of our findings is a technological artefact that facilitates the repeated and routine completion of checklists.It begins with a checklist of steps involved in machine set-up, and continues to a checklist on assessing the quality of specific parts and then moves into the care sector where it becomes a checklist of data to record about patients health and the hygiene of the care facility.At last, it becomes a checklist to assess workplace safety, again in industrial companies.
It is important to note that these travels always include a form of knowledge co-creation, in line with our epistemological assumptions, rooted e.g. in the work of [54] or [24].The artefact as it was designed for the machine-set up processes, needs to be adapted to the specific new context of application, whether it is workplace safety assessment or documentation duties in health care.This involves new knowledge about the specific context and how a tool might fit, which role it takes.This new knowledge is created between the involved actors -with or without the support of the center -in the process of appropriating the technology.While the main narrative of our findings always involves a member of the center in some capacity in this co-creative process, it does not always need to be so.The original company gave a new role to the participating employee whose duty it now is to start projects in other parts of the company, where he will be engaging in knowledge co-creation with other employees -without the engagement of the center, as far as we know so far (this might of course change in the future).Knowledge he created during his work with the center (or afterwards) will be combined with knowledge of employees about their specific challenges and practices, to appropriately address them with digital applications.
The center of competency is thus engaged in instigating and supporting the co-creation of knowledge relevant to digitization, with which participating company members might be able to support digitization within their own companies, but crucially, also to continue the co-creation of knowledge with other actors, in-or outside of their company.Sometimes, this knowledge cocreation takes the form of a specific artefact, other times it is a co-creation workshop that do not aim to develop artefacts, such as the digiXpert series.This presents an understanding of knowledge and learning as formulated by Fischer et al. [24], which state that knowledge is "mediated by artefacts, situated and distributed in a social environment" (p.3).
These types of artefact movements can be understood as a form of knowledge spillover.Although knowledge spillover is an oft-used concept in other fields, as we made clear in our discussion of related work (see, amongst others, [4,7,62,77]), it is relatively uncommon in CSCW.Hence, knowledge spillovers are carried by both the people involved as well as by the technological artefact.
The notion of knowledge spillover exhibited in our findings initially poses certain challenges for practice-oriented research and design, which is also reflected in Lucy Suchman's essay [82,83].As we outlined in the beginning, the center's work is deeply rooted in the practice-oriented paradigm of CSCW work.It usually originates in either a form of co-design workshop, such as the digiXpert series, or brief design case studies into specific organizational challenges such as machine set-up in a certain company.However, the manner in which the effects of these activities take hold in the region are by necessity outside of the specific context in which they originated and in different practice contexts.They go beyond the traditional frame of research and design initiatives under this practice paradigm.While we had intended of course for our work to reach beyond the individual companies we engage with in workshops or design projects, and tried to aid this for example via the PR work described in the previous chapter, we did not know exactly how this would take place, and the exact nature, including for example the role of mediators and multiplicators in this spillover was not anticipated.This, we argue, is however nevertheless not necessarily a departure from the epistemological orientation towards practices, but a novel extension of this approach.The spillover we observe in this study is enabled by the similarity between practices in very different contexts as well as through personal relationships between actors in the region, the centrality of which Suchman also highlights in her work [82].Knowledge is not simply 'in the air', available for everybody who is present, but spills over from one context to another via specific people, technological artefacts and activities, which is what we describe and show in this article.The challenge then becomes how to support this type of knowledge spillover from the initial site and context in which a design project was carried out.

Similarity of practices
In our example the artefact of the checklist is only able to travel from one context to another due to the relative similarities between the involved practices.Although the industrial production of car parts and the caring for patients have little similarity on the surface, both involve the similar practices of conducting and documenting small routine activities that make up the practices of setting up machines for new production processes or ensuring the well-being of the residents of care facilities.These practices are similar enough to be addressed with applications that represent digitalized checklists in which either the completion of steps or specific data can be recorded.It is this similarity that is a pre-requisite for the spillover we observe above.Even though we argue that regional work necessarily transcends individual practice contexts, practices maintain a central role in region-oriented work.Similarity of practices between otherwise very different contexts are an enabling, perhaps even necessary factor in enabling spillover.Practice-oriented research and design work which enjoys a long tradition in CSCW thus remains a crucial approach to region-oriented research and interventions.Furthermore, our findings also show a potentially fruitful relation with work on spillover in other disciplines.Particularly in the discourse in economics, the role of IT-artefacts has been widely neglected so far.
Drawing on the existing discourse on spillover introduced in section two, our observations seem to fall between the camps of MAR spillover [6,61,75] and JJ spillover [47].While the MAR concept highlights the need for geographical and contextual proximity, JJ concept highlights the need for contextual diversity, as often found in urban regions.In our case, some of the companies between which spillover occurs are contextually very different.For example, a technology that stems from industrial production is appropriated in the care sector.Nevertheless, our findings also highlight that the underlying practices are rather similar: both companies engage in forms of quality assessment, which involves routine checks which can be expressed in a form of checklist.Our findings thus construct a bridge between these two opposed positions, highlighting both the power of difference but also similarity between contexts where spillover can occur.Additionally, we also highlight the role of technological artefacts in facilitating knowledge spillover.As expressed before, the goal of the center is not to facilitate the uptake and appropriation of specific technological applications, but of the creation of knowledge and competencies within companies, with which these can undertake their own digitalization projects.The point of the spillover is thus not that the artefact finds application in two contexts, but that the associated knowledge is appropriated and applied, and further enables digitization.This confirms both the positions of Boschma & Iammarino [14], which state that a certain degree of competence overlap is a required condition for spillover, such as similar practices, as well as Bathelt [10], that the spillover of knowledge occurs only through mutual interactions between actors and face-to-face learning processes, which we call co-construction in this paper.We furthermore show that technological artefacts are a suitable vehicle for the co-construction of knowledge and thereby for knowledge spillover under the right conditions, which so far has been under-examined in the spillover discourse.This finding expands the existing discourse on knowledge spillover and points to potentially fruitful interdisciplinary research avenues within and outside of CSCW research.The focus on the similarity of practices as well as the role of technological artefacts furthermore opens the field of regional development as a field for research and intervention to praxeological CSCW work.
However, as our findings also show, similarity of practices and technological artefacts are not sufficient to initiate spillover: specific people are required to make connections from one setting to another, perceive opportunities and initiate the co-creative processes of spillover.The second part of our findings section outlined some of the activities that are required to create and maintain these connections, which we will now briefly reflect on.

Necessary work beyond research and design
As our findings make clear, the building and maintenance of networks, relationships and reputation is vital to the work of the CC.This importance is for example evidenced by the story of the employee of the regional education provider, who perceives possible synergies between her own organization's work and the center and leverages her own professional network heavily to advertise the centers activities.Something similar is observed when the member of regional administration who organized the event on the 'future of care' initiates the journey of an application from industry to the care sector through her personal networks.Also, the working group on 'knowledge management' represents a network of networks.
Personal relationships are thus a crucial component of the infrastructure for the kind of spillover we have described in the previous section.Creating, accessing and maintaining these relationships in the region becomes thus an important task if one deliberately wants to engineer for knowledge spillovers.
As a result, this becomes a continuous and crucial activity for at least some of the members of the center.Activities are carried out constantly to keep connections with partners alive and to build new ones.These activities can be very targeted and specific, such as the attendance of networking events, but also include more mundane and well-known activities such as organizing and conducting numerous meetings, email exchanges and phone calls, with the sole intention to maintain contact.The goal is either the creation or the maintenance of relationships for future collaborative activities (see experiences from other scholars, such as [3,63].
PR work is also a separate, additional component of this working towards spillover.Although this is not as directed as some other networking and relationship activities, they do involve either the maintenance of relationships to specific actors such as journalists or to mediators such as the university's press office, or the attempt to create new relationships through the preparation and publication of articles that share some of the knowledge co-created elsewhere and entice members of local SMEs to engage in further activities with the center.Not all activities, however, are included in such work packages and become thereby justifiable to the organizations involved in funding and accounting (in a financial sense) for the work.Arguably, such relationship building is important for many action-oriented research and design project and have been described elsewhere (see e.g.[3,48,63].However, they take an increased importance in this case, where regional effects are the desired outcome. These activities are not normally classified as research or design.They do, however, play a major role in such region-oriented initiatives, maybe more than in projects centered on more confined practice settings, where fewer relations are involved.This importance presents some challenges for a university-based project, as it is not a simple task to translate such activities into what is commonly referred to as research outputs, such as the preparation of research publications, or other academic activities such as teaching and other academic qualifications such as dissertations.Others [48,63], have also emphasized the need of extensive networking activities as the basis for participatory research activities, and particularly the importance of making such efforts last.This is similar to the work of [13], who emphasized the notion that 'field sites' are always constructed.While we do not intend to emphasize the constructed and hence artificial aspect of any field site, which is certainly true for our work as well, building a field site -for study and for design -is work.In the case of the CC, this means more than one field location spread across organizations and departments.This is especially important, we feel, because the regional emphasis necessitates the ongoing establishment and maintenance of connections and reputations that do not resemble 'co-design' or research.Furthermore, there might be differences between rural and urban areas.We have started this article with a brief description of the specific challenges rural areas in Germany are currently facing.We can currently only speculate on whether our activities make any difference to these specific regional challenges and more time is needed for an answer to emerge.However, there might be difference as to how network building and spillover play out in this rural region compared to an urban region.Jane Jacobs already perceived such a potential difference when highlighting the role heterogeneity plays in the creation of spillover effects and innovation.In her argument, heterogeneity within closely located organizations supports innovation and spillover.Our case somewhat deviates from this position: the region in focus is definitely rural, but nevertheless we observe spillover between rather different contexts (e.g. industry and health care).This, however, does not happen 'by itself' due to geographic proximity and serendipitous meeting of actors, but through the directed networking activities of actors in the region, in this case Ulrike, Heidi and Heike.They actively tried to build and utilize connections, create encounters as opportunities for connections and invite the expertise they deem necessary.Spillover in rural regions is thus also possible between heterogeneous organisations, but requires specific activities to take place.The networking activities described in the second part of our findings section can be understood as working towards such ends.(How exactly spillover takes places in urban regions and whether this perhaps also requires such activities is perhaps also an open questions we cannot answer).
Especially when spillover is the goal these activities take on a size and an amount of time that can not be reconciled easily with the other duties of university researchers.How to address this successfully is open to investigation.One possible avenue, we believe, is the expanded collaboration between regional actors that take on various aspects of the necessary work, including, for example, transfer organizations whose main expertise lies in the creation and maintenance of networks.

Lessons for a region-oriented CSCW
We believe the experiences and reflections presented in this essay hold several lessons for CSCW research and design work that is interested in facilitating regional effects beyond individual practice contexts.
Firstly, our findings show that regional projects are feasible, even if they come with hurdles that need to be addressed.This provides a new and exciting line of intervention for CSCW work, who usually addresses mostly single contexts.At the same time, it also opens up a possible new line of research.On the one hand, the question arises about how to conduct region-oriented work, which our account provides a first attempt at.On the other hand, we highlight the potential importance of the similarity of practices for spillover to occur.More research is necessary to understand better what nature of similarities are beneficial for knowledge spillovers.Furthermore, studying or cataloguing such similarities could be an exciting if slightly tedious beginning for an impactful regional CSCW research project, as such a catalogue would make visible the lines or routes along which the travel of artefacts and knowledge is possible.
Secondly, our study also holds more practical lessons for how to achieve regional effects.As we have made abundantly clear, relationships are crucial for spillover to occur, yet creating and maintaining such relationships represents a time investment that does not come easily for researchers whose primary activities might lie elsewhere.It has proven valuable to the members of the center under investigation to work with people who already have wide networks, instead of building them from the ground up.However, some relationship building is necessary.In our case, this is done by researcher-designers, but also by someone whose sole responsibility is public relations work.Their account makes it clear that external institutions such as the universities press office are of great help, but that this help is not sufficient, as they are not precise enough in their targeting of specific groups, but also that they do not know the project work in detail enough to initiate concrete knowledge co-creation.Perhaps then a sharing of duties is a potential way out, in which specific roles, such as a 'network manager' exist in region-oriented projects that take care of the work to make the network with key regional actors and leverage external well-positioned actors, but that are also close enough to the project to have the knowledge necessary to initiate co-creation processes or respond appropriately to external invitations.
However, there are also challenges associated to this approach, which we need to point out.We started our paper with a reflection on the challenges rural areas in Germany are currently facing.While our activities are efforts to address these specific challenges, whether our work makes a difference across the region in this regard is still an open question.This is partly due to the fact that profound regional change might require more time to be visible as such, but perhaps also a consequence of the serendipity we mentioned before.That effects of interventions travel at times via networks outside of the reach of the CC also means that we have a reduced say in where technologies and knowledge end up, to which aims they are introduced and appropriated (which is precisely part of the challenge Lucy Suchman addressed when reflecting on locating accountability).

CONCLUSION
To conclude, the experiences and reflections in this paper provide preliminary lessons on what it might take to develop CSCW projects that aim to intervene not only in specific bounded contexts but in regional innovation systems which are comprised of numerous actors, complex networks of relationships between them and a multiplicity of practice contexts.While previous studies into regional innovation systems have largely been predominantly descriptive, our study expands this discourse by investigating a culturally and geographically different region, and reflecting on efforts to intervene.Our findings show that practice-oriented approaches to CSCW research and design interventions are suitable to intervene in regional innovation systems, and offer insights into how this might be done, what challenges exist and what opportunities for (action-)research this affords.Neglecting cross-company and cross-organizational effects would completely hide an important part of the interventions' reach and impact.Regional effects, the article argues, are dependent on the travelling of artefacts from one context to another, which can be understood as a form of knowledge spillover, and which requires the co-construction of new knowledge to appropriate artefacts to new contexts.Such spillover takes place along interpersonal relationships and networks.As CSCW researchers we might not always be part of the relationships along which artefacts travel.In fact, in our findings certain external actors become instrumental in facilitating spillover through their own networks.This means we need to accept Suchman's lesson on the located accountabilities of technology designers that we have limited power to control technology use [83].Nevertheless, knowledge spillover, we argue, can be facilitated through the identification of similarity of practices between contexts, which eases appropriation and by the active construction and maintenance of relationships with actors in region.The region our efforts focus on are exemplary for the difficulties region face all over Germany.While regional development to address such issues is usually the domain of economic or political sciences, the present study opens this field up to praxeological CSCW work, by highlighting the role of practices for spillover.The activities we describe here aim at addressing some of the challenges outlined in the beginning of this paper: the workshops, events and design projects intend to enable companies to carry out digitalization activities by themselves.The digiXpert workshops train company members to take on active roles in such digitalization efforts or even to initiate them.By this they do not solely intend to create economic or financial benefit for the companies, but also save and create employment.Whether these activities and any resulting spillover create the desired change is at the current moment not answerable.Our experiences however outline a space for future CSCW studies to better understand how to carry out region-oriented interventionist projects.