Renewable energy Living Labs through the lenses of responsible innovation: building an inclusive, reflexive, and sustainable energy transition

ABSTRACT

Responsible Innovation (RI) offers Sustainability Transitions (ST) research methodologies a pathway to enable more inclusive, responsive, and sustainable transitions. Specifically, Living Labs (LLs) can benefit from RI scholarship in the scope of their sustainability research designs, as they may thereby be able to better foster critical thinking, inclusivity, and reflexivity. Drawing on a review of LLs literature and the self-reported assessments of researchers, experts and stakeholders involved in 12 LLs experiences, this research article sets out to explore how RI dimensions might serve to advance the practice of LLs in the context of ST research. The 12 LLs from seven European countries were centred on renewable energy prosumer initiatives. This analysis looks into how stakeholders were identified and involved, what inclusivity issues were addressed and what challenges were identified, to draw key lessons learned for advancing with RI in the scope of transdisciplinary energy transition studies.

KEYWORDS:

• Living Labs
• RI dimensions
• evaluation
• lessons learned
• energy transition

Introduction

The decarbonization of our global economy, ensuring a pathway towards keeping a global temperature rise at 1.5° Celsius, will require fast and transformative societal changes, and wide and inclusive participation of citizens and communities. In the scope of the energy transition, new decentralized energy systems illustrate this transformative potential as such systems imply new cultures (e.g. new shared values such as energy democracy) (Burke and Stephens 2018); new structures (e.g. distributed and digitalised energy systems) (Andoni et al. 2019), new regulatory frameworks (e.g. the European Winter Package policies) (Campos et al. 2020) and new practices (e.g. changes in energy consumption practices) (Ryghaug, Skjølsvold, and Heidenreich 2018).

This transformation would be aided by a Responsible Innovation (RI) perspective (Stilgoe, Owen, and Macnaghten 2013), as a forward-looking and ethical framework guiding experimentation and the implementation of solutions and approaches. Stilgoe and colleagues define RI as ‘taking care of the future through collective stewardship of science and innovation in the present’ (Stilgoe, Owen, and Macnaghten 2013) (p.1570). In this context and by building on the conceptual dimensions of RI further articulated by Burget and colleagues (2017), Living Labs (LLs) emerge as a forum for implementing different and interrelated conceptual dimensions (or principles) guiding such ‘taking care of the future’. These dimensions encompass inclusion (engagement of diverse stakeholder groups from the early stages of the process), anticipation (forward-looking, envisioning how the results of the research will impact the future), responsiveness (including personal responsibility for the outcomes of the research process), reflexivity (ongoing awareness of the underlining values and beliefs that are structuring the research activities), (Burget, Bardone, and Pedaste 2017; Stilgoe, Owen, and Macnaghten 2013). Burget and colleagues have also identified two additional dimensions, namely sustainability (weighing of the social and environmental impacts of the resources necessary to develop new technology and products), and care (ethical considerations about the future, and uncertainty).

In the specific context of a sustainability transition to a renewable energy sources (RES) system, and despite its undeniable importance for a decarbonised world, it has been recognised that a RES-based system may perpetuate social and environmental injustices and create new forms of exploitation and new unjust power relationships (Lennon et al. 2020; Lennon 2017). Thus, increasing attention needs to be paid to the justice aspects of this transition (Allen, Lyons, and Stephens 2019; Jenkins et al. 2020) and RI offers a pathway to facilitate social, energy and environmental justice, i.e. or ‘just transitions’ (McCauley and Heffron 2018), as well as the implementation of inclusive and reflexive methodologies fostering much needed public participation and capacity building efforts (Mormina 2019).

Living Labs have been characterized as networks or constellations of actors (Silva et al. 2019) involved in developing and co-designing new ideas, products, or services (Baccarne et al. 2016; Carayannis, Barth, and Campbell 2012), sharing knowledge, co-producing knowledge or evaluating new solutions (Mukama et al. 2022) while promoting socially responsible, reflexive, and transdisciplinary research and innovation approaches (Timmermans et al. 2020). In this context, if grounded on the dimensions of the RI framework, LLs may support ST studies (Schäpke et al. 2018).

The case of new distributed energy systems, where citizens take up an active role, as consumers, users or renewable energy (RES) prosumers (self-consuming, storing or sharing energy produced from renewable energy sources) (Horstink et al. 2020) and their ecosystems of stakeholders (Campos and Marín-González 2020), is particularly useful to assess the relevance of LLs for enacting the key RI dimensions.

However, despite the increasing number of research projects using LLs (Timmermans et al. 2020) and their high potential for illuminating the different complexities and dimensions of sustainability problems (Baran and Berkowicz 2020; Voytenko et al. 2016), insights into how LLs can contribute to a more inclusive transition and foster RI uptake are still piecemeal, specifically within the scope of the energy transition and energy justice scholarship – a gap which this article aims to address.

Living Lab approaches can facilitate the uptake of RI as they provide a space or forum for experimenting and for reflecting proactively, as new solutions and approaches emerge. This is important for gaining insights into RI while promoting RI through multi-stakeholder collaborations in practice (Timmermans et al. 2020). Despite having a clear orientation towards inclusivity, evidence has shown that LLs can equally lead to reinforcing pre-existing exclusionary practices within a specific context (Evans and Karvonen 2014). Thus, this paper aims to utilise perspectives of RI to advance thinking and practice in LLs. It does so by drawing on an inquiry into how LLs apply the conceptual dimensions defined by RI, and by looking into the use of LLs in the specific context of the energy transition.

Living Labs have the potential of bringing to the foreground the ethical importance of responsibility, considering the future impacts of today’s technological and social change, and the need for fostering inclusive and reflexive participatory processes that can trigger advancements in social, economic and environmental sustainability (Adam and Groves 2011). As an open space for discussion, LLs can equally offer a means to deliberate on different visions for a sustainable future (e.g. between economic green growth and degrowth (Nelson and Edwards 2020)), and address underlining differences in value perceptions (Martin, Evans, and Karvonen 2018). For instance, considering the different geographies of renewable energy, a large-scale solar or wind energy project may be considered unsustainable by some, due to its inevitable environmental impacts on local biodiversity, while others would still support such projects based on the need for a fast decarbonisation (Fast 2013). LLs can equally offer the grounds to openly discuss and find trade-offs on different value perceptions of what sustainability means, including economic externalities of renewables (e.g. the need for extensive mining) and the social impacts of large-scale installations for local populations and their livelihoods (Calvert 2016). As the energy transition picks us speed, research is needed that investigates how these complex problems can be addressed by LLs in order to better understand their potential to offer a forum for discussing different value perceptions and trade-offs, across multiple stakeholder interests.

Thus, the driving research question of the article is to explore how to utilise RI dimensions and related perspectives on inclusivity to advance the practice of LLs in the context of ST research. Taking stock of a thematic analysis of the reported assessments of researchers, experts and stakeholders involved in 12 LLs experiences, from 7 European countries, focusing on RES prosumer and citizen-centred energy systems, this article aims to contribute to consolidating the inclusion, anticipation, responsiveness, reflexivity, care, and sustainability dimensions of RI, by offering a set of lessons learned from practice.

The paper is organized as follows: definitions and key concepts of Living Labs and their relevance for RI are the focus of the next section. Next, the methodology is detailed considering key research questions, data sources and an analytical strategy. Results are then provided in relation to the research questions and discussed in view of RI literature. Final conclusions reflect on how Living Labs can contribute to RI in the scope of the energy transition.

Living Labs as spaces to investigate the uptake of responsible innovation

Definitions of Living Labs

Living Labs have been adopted in the scope of a diversity of studies on new technologies (Mukama et al. 2022), climate change policies (Baccarne et al. 2016), food production and urban development (Voytenko et al. 2016), or low-carbon economy (Baran and Berkowicz 2020), to cite a few examples. Methodologically, LLs typically rely on multi-method approaches (e.g. workshops, focus groups, interviews). This diversity is also present in the contexts in which LLs have been implemented (e.g. research, technology assessment, policymaking, education). However, a crosscutting aspect to all LLs approaches is the relevance of a real-life space and the need for an interactive and participatory approach.

Multiple definitions and understandings exist of what a LL approach entails – e.g. ‘urban Living Labs’ (Voytenko et al. 2016), ‘real-world laboratories’ (Renn 2018), ‘social labs’ (Timmermans et al. 2020) and ‘policy labs’ (Lewis 2021; Olejniczak et al. 2020). This diversity indicates that LLs can be applied in different contexts, from the development of new products and innovations to the design and implementation of new policies.

Voytenko et al. (2016) consider ‘urban LLs’ as a ‘form of collective urban governance to produce innovative solutions’. These authors also distinguish three characteristics of LLs, namely geographical embeddedness (they are situated in a specific space); experimentation and learning (they are primarily driven by the challenges posed by complex problems); and participation and user involvement (they offer a new platform for participation that should be openly available to the many).

The ‘real-world laboratories’ literature highlights the potential of LLs as spaces for transformation and reflexive learning and as ‘large-scale research infrastructures’, which are not circumscribed to specific interventions in time, but rather establish an ongoing space for reflexivity (Engels and Walz 2018). Within this research framing, LLs have emerged in different scales and geographies, such as households, cities or rural regions (Brons et al. 2022).

In RI literature, ‘social labs’ are defined by Hassan (2014) as ‘platforms for addressing complex social challenges’ (p.3). Further elaborating on this concept, Timmermans and colleagues (2020) distinguish social labs from LLs, depicting them as more inclusive ‘spaces for experimentation’, with the ‘active participation of a range of societal stakeholders’, following a ‘multi – and interdisciplinary expertise’ and supporting ‘solutions and prototypes on a systemic level’ (p.414-415).

‘Policy labs’ are also a type of LL, where the focus is on developing new policy and democratic innovations, and are often set up by policymakers, governments and/or civil society bodies (Lewis 2021).

In synthesis, conceptualisations of LLs as ‘innovation space’ are very much in line with policy labs focusing on designing democratic innovations (Olejniczak et al. 2020), and with user experience design approaches (Pallot, Pawar, and Santoro 2013). Yet, drawing on the participation of multiple stakeholder groups around a specific topic, LLs experiences become increasingly interactive (Voytenko et al. 2016). With a focus on co-creating new designs for a more sustainable future, LLs have also been proposed as a large scale infrastructure that enables effective transformation research (Schneidewind et al. 2018). Furthermore, while bringing the focus onto human agency, social labs (Franz 2015; Timmermans et al. 2020) introduce LLs as spaces for fostering reflexivity and co-learning with a focus on social change. Figure 1 illustrates these defining elements of LLs that are identified in the literature.

Figure 1. Synthesis of the different defining dimensions of the living labs approach.

In this study, LLs are understood as real-life settings (in which people may engage physically or virtually) for responsible reflexivity, co-learning, and the co-production of knowledge, involving a group of actors who come together to collectively co-produce and reflect on the planning, implementation and/or assessment of new ideas, solutions, products, or services. The LLs were implemented by a team of researchers, which included the two authors (who co-organized with other colleagues the implementation of the Portuguese Living Labs). The team of researchers engaged several stakeholder groups in the activities of the different Living Labs implemented in a total of seven European countries.

Inclusivity in Living Labs

If LLs are spaces for innovation and for the co-design of new policies, inclusivity should be central to enable the effective development of products and services that are able to respond to user needs (Franz, Tausz, and Thiel 2015). However, it has been found that innovation-centred approaches may fail to properly engage local communities and integrate local needs and specificities, when their focus is mainly on demonstrating and testing new technologies and services (Hodson and Marvin 2009). Such approaches may equally reinforce existing exclusionary practices in public and citizens’ participation, as was concluded by Evans and Karvonen’s analysis (2014) of a low-carbon urban laboratory in the UK.

As spaces for reflexivity and transformation, Schneidewind and colleagues’ review (2018) shows how LLs have been adopted within different ‘schools’, although it does not specifically address inclusion as a principle for the variety of research fields where LLs have been used. Thus, the Transdisciplinarity approach (Lang et al. 2012), understands LLs as an element of wider transdisciplinary processes, while the Transition Management approach recognizes LLs as spaces for innovation and learning (Nevens et al., 2013) and the Transformative science approach deems LLs as ‘reflexive learning spaces, which combine practice and scientific approaches’ (Schneidewind et al. 2018) (p.13). In this space of learning, experts, civil society actors and other stakeholders cooperate in the co-coproduction of knowledge for co-creating more sustainable societies, yet whether a wide representation of citizens is included, and if so how, is not examined.

Intervention focus of Living Labs

LLs are an important topic of interest in innovation studies (Schuurman 2015). In this context, LLs have been characterised as ‘ecosystems’ where end-users and other stakeholders are involved in the development of innovation over a significant period (Baccarne et al. 2016). Market innovation approaches highlight the relevance of LLs as ‘testbeds’, but also as ‘collaborative working environments’ (Niitamo et al. 2006). According to Niitamo et al. (2006), LLs can be defined as: ‘an emerging public-private partnership concept in which firms, public authorities and citizens work together to create, prototype, validate and test new services, businesses’ (p.2). Of note is the vagueness of Niitamo’s definition of public authorities in this context and whether this stakeholder group includes universities and administrators or merely policy actors. Ballon and Schuurman (2015) introduce a structure for a clearer stakeholder definition, by proposing the use of methodological analytical levels. The organisational (macro) level is where the stakeholders are identified, and where an engagement process is initiated and developed. The managerial (meso) level is where the LLs’ specific topics, working plans, expectations, and goals are collectively managed. Lastly, the practice (micro) level refers to the specific methods applied in each intervention (or activity).

Nevertheless, the innovation-centred approach of LLs places a strong emphasis on how new technologies are experienced by users (Ballon and Schuurman 2015; Schuurman 2015). Thus, these applications of the LL approach have resulted in shorter-term interventions, involving specific ‘mini publics’ (Setälä 2017) namely those framed by the usability of the innovation, solution or product being co-designed (Voytenko et al. 2016). More recently, an evaluation of ‘policy labs’ (Lewis 2021) – another example of LLs as spaces for innovation – , highlighted as a key barrier the short life of these processes, particularly without appropriate leadership on the part of those responsible for overseeing activities. By primarily focusing on short-term processes rather than multiple interventions within a longer timeframe, the continuous development and implementation of innovations can be compromised (Hartley et al. 2019). Furthermore, by concentrating solely on usability issues, aspects such as future uncertainty regarding technology innovations and their impacts on new social practices, as well as broader unforeseen societal implications can be overlooked, thus compromising the sustainability and care dimensions of RI (Sovacool, Hess, and Cantoni 2021).

Responsible innovation scholarship and guiding research questions

Studies relevant for inclusivity in RI published in this journal have highlighted the challenge of accounting for a diversity of publics, interests and needs across increasingly complex societies (e.g. Brand and Blok 2019; Hartley et al. 2019; Nelson, Selin, and Scott 2021). This has been termed a ‘societal alignment dilemma’ (Ribeiro et al. 2018), emerging from the need to cater to different publics, societal needs and diverging interests across stakeholders and communities. This dilemma is crosscutting to all dimensions of RI, as it requires considering any unwarranted impacts of envisioned technologies, while accounting for the different values, needs and ideas about the future of diverse publics. Realistically, it becomes necessary to accept there are limitations to societal alignment, as stakeholder biases and institutional constraints are likely enduring and often barriers in RI (Kuzma and Roberts 2018). For instance, the LLs analysed here were driven by the need to investigate the social aspects of the energy transition and were built on the (changing) interactive needs and aspirations of the participating stakeholders. The resulting social innovations, such as energy communities, are dependent on the interest and trust of local communities to participate, on the availability of new technological structures, such as smart meters and distributed energy systems, and on new institutional frameworks, such as new energy legislation (Gjorgievski, Cundeva, and Georghiou 2021), thus posing critical challenges for pursuing social alignment.

In this context, our inquiry considers how LLs can offer a forum to explore in practice ecologically driven and community centred innovations that foster environmental, social, and economic sustainability, while pursuing inclusivity and social alignment. This study equally aims to assess the relevance of the key dimensions of RI, namely inclusion, anticipation, responsiveness, reflexivity, care and sustainability (Burget, Bardone, and Pedaste 2017; Stilgoe, Owen, and Macnaghten 2013) in the practice of LLs. These dimensions inform our three specific research questions:

• RQ1. How do the LLs describe and identify stakeholders for participation?

• RQ2. What inclusivity issues are addressed?

• RQ3. What challenges did participants and researchers identify?

Considering the relevance of the last question for RI, literature has emphasized the need to be responsive to societal challenges (Stilgoe, Owen, and Macnaghten 2013), while also dealing with context-specific challenges that may emerge, for instance, when applying RI to controversial technologies as in, e.g. the agricultural sector (Gremmen, Blok, and Bovenkerk 2019).

Methodology

Data and sources

The design and implementation of this research received an ethical approval from the ethical commission of the Faculdade de Ciências da Universidade de Lisboa, which was also signed by all the research teams involved in the project and submitted to the funding institution prior to advancing with the study.

The methodology draws on a qualitative evaluation of 12 LLs implemented in Belgium, Croatia, Germany, Italy, The Netherlands, Portugal, and the United Kingdom in the context of a European funded research and innovation project. These countries offered examples of different degrees of citizen participation in decentralised energy systems (Horstink et al. 2020), and therefore provided a starting point to develop and compare LLs experiments in the context of prosumerism and the energy transition. The 12 LLs all aimed to seek out new solutions, products, or services for helping RES prosumers, stakeholders and citizen groups involved in developing new prosumer projects, and overcome legal, technological, financial and/or cultural barriers (see Table 1).

Table 1. 12 Living Labs (LLs) in Europe, main goals, and their achievements.

LL Code	Country	LL name	Main goal of the LL	Extent to which the goal was achieved
BE	Belgium	Getesnippers Living Lab	Evaluate the feasibility of a valorisation chain for sustainable heat using residue biomass from landscape conversion and maintenance work.	Only partially, as some stakeholders were not successfully involved, and more information is needed to advance with a robust regional plan.
HR1	Croatia	Island of Silba Living Lab	Assess the feasibility of clean water supply through desalinisation and PV systems involving the local community.	The goal was achieved. The LL resulted in a feasibility study, involving the local community and plan for a new energy-water system
HR2	Croatia	One-stop-shop Living Lab	Develop a tool to serve prosumers in the future by showing the bankability of individual PV installations.	The goal was achieved. The living resulted in a new tool that enables prosumers to have a clear understanding of their future investments in green energy. Yet, the tool could be more intuitive and user-friendly.
DE1	Germany	Northeast Lower Saxony Living Lab	Evaluate the possibilities to market a regional electricity product and widen the citizen-led projects outreach.	The goal was not achieved. The cooperatives involved found they still need to learn more about regional marketing.
DE2	Germany	Sub-Urban Heat Transition Living Lab	Development of a local district heating network and evaluating the extent to which residents, and a local school are willing to engage with the project.	The goal was achieved. It was possible to involve the local school in the district heating project, but doubts remain on whether local actors will be able to manage the implementation of the project.
IT	Italy	The Santorso Living Lab	Provide citizens with support to enlarge their scope of action about RES (biomasses, sustainable electric mobility).	The goal was achieved. A new ‘help desk’ was created for citizens. Also, as a result, the municipality is developing a new renewable energy community.
NL1	The Netherlands	Buurtwarmte Living Lab	Support the development of neighbourhood energy cooperatives providing sustainable heating to their members.	The goal was not fully achieved. Yet, LLs members found they had set too high expectations for the process, within a two-year timeframe.
NL2	The Netherlands	Aardehuis Living Lab	Find the institutional barriers around prosumerism in the eco-village and co-create solutions together with relevant stakeholders to overcome obstacles.	The goal was achieved. The LL enabled a ‘deepening of knowledge’ in different areas (i.e. regulation, ecological footprints, etc.) targeted by the activities.
PT1	Portugal	Wines of Alentejo Living Lab	Support the adoption of RES by the Alentejo’s viticulture industry by proving information and promoting the setup of collective self-consumption schemes.	The goal was achieved. Wine producers in the region were effectively involved and a new design for a local renewable energy community, as well as for new partnerships was developed.
PT2	Portugal	Sao Luís Energy Community Living Lab	Increase energy literacy and facilitate the setup of an energy community in Odemira, Alentejo.	The goal was achieved. A new renewable energy community is developing.
UK1	United Kingdom	Bristol Energy Cooperative Living Lab	Explore new potential business models to guarantee further exploitation of large-scale renewable energy plants even without a fixed remuneration guaranteed under the previous FiT system.	The goal was not achieved. The LL was found to be helpful in ‘highlighting key issues, but not was not able to solve them.’ COVID-19 lockdowns were a key barrier.
UK2	United Kingdom	Bristol Energy Company Living Lab	Investigate how emerging business models, such as ‘heat as a service’ could contribute to the UK’s net-zero ambitions.	The goal was not achieved. COVID-19 lockdowns and economic downturn were key barriers.

The LLs followed a three-stage approach, from exploration to implementation and evaluation (Schuurman 2015).

While the exploration stage implied an assessment of key barriers and challenges that were relevant for LLs members, the implementation stage included various interventions through which these challenges were addressed. Such interventions were implemented between September 2018 and September 2020. The LLs involved RES prosumers and their ecosystems of stakeholders, i.e. market representatives, policy actors, civil society, and community-based spheres, at different governance levels (i.e. local, regional, and national). The main stakeholders that collaborated directly with the researchers throughout the LLs are referred to as ‘LLs members’ and make up a total of 30 people.

A non-hierarchical structure was used during the exploration stage to ensure that all stakeholders involved in the LL had equal opportunities for participation. This organizational feature of the LLs studied allowed us to share plans and co-design the sessions with the LL members (e.g. agendas, location, definition of objectives and activities) resulting in ‘operational roadmaps’ that guided all the interventions during the implementation stage.

Circa 500 people participated in a total of 48 interventions across the 12 LLs. Some results of specific LLs have been published in (Brown, Hall, and Davis 2019; Campos et al. 2019; Pontes Luz and Amaro e Silva 2021) and the overall results informed the analysis of RES prosumerism in (Wittmayer et al. 2021).

Lastly, the evaluation stage was based on a final self-assessment of researchers and LL members, and stakeholder interviews, resulting in 12 evaluation reports. Table 1 summarises the main challenges that the LLs sought to address, and to what extent the different goals were achieved. It also indicates the name codes attributed to the LLs (e.g. BE, HR1, etc.) which will henceforth be referred to by these codes. In addition, Table 2 shows the different stakeholder types and number of interviewees, which provided inputs to the evaluation reports.

Table 2. Stakeholders interviewed in the 12 LLs.

Stakeholder type	Description	Number of interviewees and respective LL
Research	Researchers involved in the LLs research (design and implementation of LLs activities)	14 researchers and experts from the 6 research institutions and two civil society organisations involved in the design and set up of the 12 LLs
LL members	Representatives of municipal agencies	9 (2 in BE, 3 in DE2, 3 in IT, and 1 in UK2)
LL members	Citizens (e.g. neighbourhood associations, members of local council, schoolteachers, and students)	10 (5 in HR1 and 2 in PT2, 3 DE2)
LL members	Representatives of energy cooperatives	5 (2 in DE1, 2 in PT1 and 1 in UK1)
LL members	Representatives of energy businesses and other companies	4 (2 in NL1 and 2 in PT2)
LL members	Informal community (eco-village)	2 in NL2

Analytical strategy

The methods used for the LL qualitative data collection included semi-structured interviews (conducted in-person) and self-administered questionnaires (applied online) during or after the last workshops implemented in the 12 LLs. A total of 30 interviews were done with LLs members and 14 questionnaires were filled out by the researchers. The questions included covered diverse topics related to the LLs activities, yet this study focuses only on the questions that are relevant for addressing the three research questions. Table 3 shows the interview questions made to both LL members and the researchers, in relation to the research questions they sought to address.

Table 3. Questions addressed to LL members and researchers involved in the LL.

Research Questions	Questions directed to specific participants
RQ.1 Stakeholders	How do you think the LLs benefitted the stakeholders involved? (researchers) Do you think you managed to co-create the interventions with stakeholders? Why and if yes how? (researchers) Did you feel you had a role in co-creating the LLs activities? How much did you feel involved? (LL members)
RQ.2 Inclusivity	What is the main lesson learned from all the interventions? (i.e. if the LLs process had been a story, what key lesson would you take from it?) (LL members) How do you think the LL benefitted the stakeholders involved? (researchers) 6. Do you think the LLs involved the right people? Were they inclusive? If not, who do you think was missing and why? (researchers)
RQ.3 Challenges	Following the activities in which you participated, what do you think are the most important strengths and weakness of the LLs? (LL members) How did the LLs help solve the needs/challenges identified at the beginning? (LL members) 9. What suggestions do you have for future research with LLs (e.g. management strategies, methods, communication, etc.)? (researchers)

Data from the in-person interviews (transcripts) and self-administered questionnaires (text files) were collected in 12 evaluation reports, one for each of the 12 LL, and delivered by the researchers responsible for their implementation. These reports contained additional accounts of the overall process of the LLs, including the results of the workshops and of the specific interventions implemented. Although the analysis of specific outcomes of each living lab is beyond the scope of this article, the evaluation reports, which are a compilation of the evaluation data collected from each LL, offered key insights for a comparison of the overall direction that each LL took.

To derive the final lessons learned from LL practices and address these research questions, data analysis included thematic coding (Braun and Clarke 2006) and an interpretivist stance (McChesney and Aldridge 2019). An inductive approach to thematic analysis (Braun and Clarke 2012) was found useful to identify key themes that allowed addressing how a LLs approach can help to promote inclusiveness, anticipation, reflexivity, responsiveness and sustainability in energy transition-related processes.

After a close familiarisation with the data (interview transcripts and replies from the self-administered questionnaires, integrated into the 12 evaluation reports), a first round of inductive coding generated a set of initial codes that provided relevant insights to be collated into potential themes. This phase of sorting and collating codes into themes was followed by reviewing, defining, and naming themes, which resulted in a collection of final themes that worked in relation to the entire data set and that contain the ‘essence’ of the data captured by each of them. To safeguard validity, the final codes and themes were then compared to the original dataset, reverting the process, and supporting the final extraction of lessons learned. This process of data analysis was interactive, as both authors of this paper read all materials and were involved in the entire coding of the data, as well as in subsequent analysis. All data was hand-coded, and the authors re-analysed each other’s codes, resulting themes, and lessons learned, before agreeing on a final version.

Researchers and experts as well as LLs members were asked to offer insights into key lessons learned from their experience (see Table 3, question 4). Thus, after coding their replies and considering the overall results of the thematic coding, the final lessons learned are further shaped by an interpretivist stance (McChesney and Aldridge 2019), which enabled looking for emergent codes that directly result from the data collection, while considering the context of the research – i.e. the specific story of each LL. The lessons learned are thus a direct result of the thematic coding of the responses to the questions relative to the lessons learned, and of a second level of analysis, which consisted in confronting the thematic coding results with the development of each LL and their reporting on the key ideas, services, plans and projects that resulted from the interventions implemented.

Results

The thematic analysis of the 12 evaluation reports resulted in the generation of 40 codes – i.e. 22 from the LL members and 18 from the researchers– that turned into crosscutting and recurrent themes, such as ‘knowledge’, ‘co-creation’, ‘stakeholder engagement and participation’, ‘inclusivity’ and ‘limitations and challenges’. In what follows the results (i.e. codes, themes and lessons learned, with examples of text extracts) are presented in relation to the three research questions, and to the dimensions of RI (i.e. inclusiveness, anticipation, reflexivity, sustainability, and responsiveness).

Stakeholders – how do the Living Labs describe and identify stakeholders for participation?

Stakeholders were identified across all LLs based on the local context and the specific goals of the LLs. Due to the wide variety of contexts, stakeholders included representatives of municipalities (in all cases, except HR2, NL1 and NL2), local Parish administrations and citizens (in HR1, PT2, IT, NL2), technicians, and technology and economy experts (in all LLs), large energy companies (energy transmission operator and energy distribution operator in UK2), energy cooperatives, start-ups and SMEs involved or not in energy systems (in BE, DE1, DE2, HR2, NL1, PT1 and PT2), civil society organisations (in all cases), schoolteachers and students (in DE2), and eco-villages (in PT2 and NL2).

‘Stakeholder’ was a recurrent code across the data, as all LLs described themselves in terms of the different stakeholder groups involved and their relevance for advancing with the LL activities. LL members themselves identified the key stakeholders they considered more important for the development of the process. Furthermore, stakeholders changed over time according to the needs of the process. For instance, BE was guided by the goal of ‘creating a business case for sustainable heat from wood chips’, yet acceptance issues from the side of citizens led to deviating from the initial goal and instead focusing on bringing different stakeholders, including citizens, into the conversation. Researchers from the same LL concluded the process ‘enabled stakeholders to approach their work in a holistic and more structured manner’. For PT2, it was considered a good practice to collaborate and involve stakeholders early on. As stated by one of the researchers involved in the LL ‘we involved our representatives in the preparation of the workshops, from setting together the goals to who to invite (…). We developed all the interventions together. Even the last one, which was online (due to the pandemic situation) was still co-developed with participants (…). A member of the community was responsible for graphical support’. Similarly, in PT1, the establishment of collaborations offered a pathway forward, as put by a researcher ‘(…) when encouraged to do so, business competitors in this (wine) sector can find ways to collaborate well and plan projects together, (…) they had to work together to scale up the adoption of renewable energy in the region’.

‘Empowerment’ was also a recurrent code, and in all LLs, participants and expert teams agreed that the process was empowering. As noted by a LL participant from IT, ‘we also gained confidence in ourselves. Looking at other experiences, we realised that we have the competences, local knowledge, and governance structure to be able to become a better prosumer community’.

The importance of ‘trust’ was equally emphasised in relation to the stakeholders involved. LLs were understood as places that could build up trust between local communities and multiple stakeholder groups. For instance, a participant in PT2 found it important to ‘trust the collective creativity of the group’ (…) ‘new solutions emerge from this collective co-creation process, even if it takes more time’ (…). Next to trust, ‘working together’ and ‘collaboration’ were recurrent codes, as put by one research team member from PT2, ‘LL partners (…) were constantly being involved at the different stages of the work developed (i.e. including in co-designing the interventions, etc.), they felt they were working truly together with the research team, rather than being led by the team’. Other important codes included the relevance of ‘communication’ and ‘networking’, as LLs were also understood as places for establishing new networks and interrelations, and this was perceived by team members as a critical for success in engaging multiple stakeholders.

A key lesson learned for research teams was the importance to have a clear engagement strategy that considered different incentives (e.g. networking opportunities, lobbying) to ensure a high enthusiasm and the participation of multiple stakeholders and citizens throughout the process (i.e. in the case of HR1, HR2, NL1, DE2).

‘Sustainability’ was important for the LLs process. Both research teams and LL members found that the new ‘collaborations’ and ‘networking’ activities that were established also contributed to accelerate efforts towards achieving sustainable energy transition-related goals. For instance, at least in two LLs, new energy communities were being developed because of the LL activities. Except for UK1 and UK2 LLs, all others benefited from the co-creation of tangible tools which accelerated local efforts towards transition goals. The two exceptions in the UK were also due to structural constraints beyond the scope of the LL’s activities (e.g. bankruptcy of the prosumer projects that were being addressed by the LL).

Table 3 summarises the key codes related to the stakeholder’s theme, and the final lessons learned drawn from the interviews to participants and research teams.

Inclusivity – what inclusivity issues do the Living Labs address?

Inclusivity was mainly understood in terms of the variety of stakeholders involved. Only two LLs (i.e. NL2 and PT2) indicated a direct concern with inclusivity in terms of gender, age, educational, ethnic, or socioeconomic aspects. According to a LL member from NL2 ‘the gender diversity of the workshops was rather inclusive, as the living lab included a broad range of female and male participants. However, it lacked cultural diversity.’ The person also claimed that this lack of diversity ‘reflects the diversity of citizens living in the community where the living lab was set’. Research teams involved in IT pointed out that ‘it was inclusive having sit around the table not only representatives of the municipalities but also technicians and some volunteers of the local group of citizens. Given the nature of the LL (very informative) the participants pointed out that the activities could have been opened to all the citizens’.

Thus, when asked about whether the LL was inclusive, key codes in the responses were ‘stakeholder’ and ‘representativeness’. Yet, an inclusive LL was understood by most participants and research teams as one where a wide diversity of stakeholders was represented, but not in terms of citizen (gender, socioeconomic, or ethnical) representativeness. For instance, HR1, IT, PT2, and UK1 specifically addressed the issue of energy poverty, yet, except for HR1, all recognised they did not directly involve more vulnerable citizens in their meetings and workshops. Thus, ‘representation’, by e.g. involving local civil society organisations or representatives of the municipality, was the main LL strategy for addressing inclusivity.

Responses to the lessons learned also referred to representation. According to a participant from PT2, ‘the main lesson learned is that by just bringing different people, with different interests and visions into the same room, to focus on a specific topic (i.e. increasing local decentralised energy production), we (…) help advance further with the transition. So, moving beyond the local community work, it is important to join different types of communities, with different experiences (…) different interests (…); (…) different visions for the transition (…)’. When considering the process retrospectively, HR1 research teams found that ‘the main lesson learned is that inhabitants and the local community are the ones who should be included, informed and a part of every big and important project such as this one. If the local community is not aware (…), there is no interest from the authorities to make their lives easier’. The importance of simplifying languages was also highlighted. For instance, IT participants found that ‘a higher participation could have been better, and we will work to manage meetings to enable the messages we deliver to be simplified and as less ‘technical’ as possible’. In this example, the need to ensure further attention to inclusivity appears to have been a lesson learned by LL participants as a result of the interventions.

Not attending to inclusivity was a shortcoming of the LLs’ design and a key lesson learned, as overall LLs did not depart from a clear definition of inclusivity that ensured gender, ethnic, socioeconomic, and cultural diversity amongst participants. Research teams also did not pay sufficient attention to gender and cultural diversity, diversity in age and socioeconomic groups, although equal access to information was a concern, as well as attention to potential digital gaps among participants (e.g. in IT and DE2).

Table 4 summarises the key codes related to the inclusion theme, and the final lessons learned based on the accounts of participants and research teams.

Table 4. Stakeholders: Codes and lessons learned.

Codes	Trust Diversity Experiences Interests Knowledge Solutions Communication Networking Common goals Participation Local Empowering Collaborations Sustainability
Lessons Learned	− Stakeholders were empowered and gained a sense of working towards a shared goal. − Communication, awareness raising, and networking activities were crucial to engage people with different interests. − Stakeholders’ identification and involvement needs to be context driven.

Challenges – what challenges did participants and researchers identify?

A key code related to challenges was the ‘capacity’ of both researchers and participants to respond effectively to the needs of the LLs. For instance, in NL1, researchers found it challenging ‘to find an intervention that is useful for both researchers and the participants’. Capacity also related with local contexts and the ‘time’ available to deepen knowledge and information sharing processes. For instance, in NL2, a participant notes she ‘would have liked to have gotten more on the content. This has to do with the deepening of knowledge […] I would have liked to get more out of it, […] extract more knowledge out of it’. DE1 researchers also found that it ‘takes a lot of working hours to accompany such projects from idea to implementation. This time is not always available to participants and researchers.’

Another crucial factor was ‘lack of engagement’ of stakeholders on the topic of prosumerism and renewable energy production, which was, according to HR2, an important lesson learned, i.e. – ‘the market is aware of the benefits of the prosumer solutions, but there is still a missing link from that awareness to the realisation and larger scale up of the projects being implemented’. For IT, DE1 and DE2, engagement was particularly challenging at the beginning stages. A DE2 participant found ‘difficulties in the acquisition of participants and feedback’ and for DE1’s researchers, ‘citizens were really hard to engage in the workshops’, and engagement was characterised as ‘a struggle’. In HR1, a major challenge was the ‘low interest of certain stakeholders (regional governance and local water supply company) to participate in the project’. The pandemic situation was pointed out as a key challenge for sustaining engagement. Also, the pandemic evidenced pre-existing digital divides and thus conditioned the participation of some LL members, when in-person events were converted into online formats. For instance, in DE1, researchers found that ‘many people of the LL were not familiar with basic computer skills’.

‘Impact’ after the end of the LLs was also a key code for this topic. In HR2, researchers found that ‘the main weakness is that, after the living lab is done, there are few instruments to further continue the projects’. Although in the case of HR1, participants acknowledged that the process will ‘continue (…), led by a local cooperative and NGO Island Movement.’

Another challenge pointed out by participants in PT1 and IT was coded as ‘distance’. For instance, in IT a participant found that ‘the institutional nature of part of the LL makes it slightly ‘distant’ from grassroots initiatives.’ Also, as put by a BE participant ‘(…) some of the best practice examples (from other European countries) were too far from our own context.’

To include moments for ‘reflexivity’ was found to be both challenging and important, as pointed out by NL1 and NL2 LL members, who agreed that on ‘hindsight it would have been nice to have a reflection of the process (…) set-up in a more structural manner’. In addition, BE researchers found that the management of expectations was crucial as ‘the reality of the LL does not always live up to the ideal of co-creation’. Reflexivity (in the sense of the reflexive awareness of the researcher) was itself a lesson learned, as LL members and research teams recognised the relevance of integrating additional moments for ‘sensing, listening, co-learning’ (NL1, PT1, IT).

Another key challenge for BE, DE1 and NL2 was to navigate across highly ‘heterogenous’ groups. Researchers from DE1, for instance, found that the ‘heterogeneity of knowledge between the participants was challenging because inputs needed to simultaneously address people with almost no knowledge, as well as people with deep knowledge on the topic’.

Living Labs experiences helped participants gain clarity about their shared goals and ambitions. For instance, IT1, PT1 and PT2 stakeholders explicitly manifested how new possibilities for the future became clearer. Yet, a key challenge identified in envisioning the future energy system (i.e. the anticipation principle), related to different levels of knowledge about the transition. While some participants were very well informed of technical, economic, or legal aspects of renewables, others struggled to get a hold of the new knowledge (e.g. PT1, NL1, IT). Thus, the LL interventions were marked by a concern with sharpening participants’ goals about the future. NL1, PT1, PT2, DE1, DE2 and IT, all included presentations of invited expert stakeholders to help level the knowledge and understanding of shared goals (Table 5).

Table 5. Inclusivity: Codes and lessons learned.

Codes	Gender Cultural diversity Involvement Messages Representativeness Gender diversity Cultural diversity Inclusivity
Lessons Learned	Inclusivity was mainly considered in relation to the diversity of the stakeholder groups involved in the activities and the knowledge and expertise they offered. − Gender and cultural diversity were not central concerns. − LLs require attention to inclusivity, in terms of the diversity of citizens and stakeholders involved.

‘Responsiveness’ equally relates to the main lessons learned, regarding challenges. For instance, LL members found they should agree early on their expectations for the results of the LLs. To achieve this, guiding dimensions, such as co-creation, should be defined at an operational level as an overarching concept to guide all activities. The management of expectations was considered a critical and challenging aspect, requiring clarity and openness to changes in the goals and direction of the LL. In addition, ensuring financial compensation for all LL members (not just research teams) could have helped deal with unexpected external challenges, included those posed by the COVID-19 pandemic (Table 6).

Table 6. Challenges: codes and lessons learned.

Codes	Engage Capacity Impact Distance Reflexivity Heterogeneity Local contexts Pandemic situation/digital divides Responsiveness Anticipation
Lessons Learned	− More co-creation and higher reflexivity and monitoring were missing. − It is critical to be prepared for sharing knowledge effectively among heterogenous groups of participants. − It is important to set up a strong engagement strategy with attention to local contexts and dynamics.

Discussion

Out of the 12 LLs, only seven managed to achieve the goals they set out to pursue. Nevertheless, all were successful in accomplishing some degree of knowledge exchange, and in advancing (even if only partially) towards reaching their energy transition-related goals. The analysis of their experiences offered insights into how stakeholders were identified and characterised, about their perceptions regarding inclusivity and about the challenges met throughout the process, resulting in key lessons learned that can be useful for the integration of RI dimensions in LLs (Burget, Bardone, and Pedaste 2017).

The stakeholders involved consisted of different actors across the market, civil society, community, and policy spheres, with a diversity of knowledges, competences, and capacities. This multi-stakeholder perspective was mainly informed by conceptual frameworks such as the quadruple and/or quintuple helix approaches (e.g. the involvement of universities, governments, industry, civil society and environmental organizations) (Baccarne et al. 2016; Carayannis, Barth, and Campbell 2012). This is consistent with sustainability transition research’ understanding that innovations in transitions are co-produced by multiple social actors with different institutional logics (Avelino and Wittmayer 2016). Less attention, however, was paid to involving a representative and diverse group of citizens (Bifulco 2013). Thus, societal alignment efforts were not easily met (Kuzma and Roberts 2018).

In the 12 LLs, concerns such as gender balance, and the inclusion of citizens representing various ethnic, age, and cultural groups, with diverse values and needs, were for the most part overlooked. Another limitation was that the majority of those who benefited from the LL process were already actively involved in local energy transition projects. Thus, the need to ensure representativeness of those most often excluded from innovation and co-creation processes, which is fundamental for RI and social labs (Franz 2015), was not fulfilled. As spaces for innovation that can inform the implementation of new green technologies and energy transition policies (Engels, Wentland, and Pfotenhauer 2019) LL structures that ensure the representativeness of all citizen and stakeholder groups who may be either losers or winners of the outcomes of transition processes, need to be further developed.

Responsible Innovation is a forward-looking and anticipatory approach (Adam and Groves 2011; Stilgoe, Owen, and Macnaghten 2013), likewise the 12 LLs informing this paper’s analysis were focussed on the future – e.g. the transition to decentralised sustainable energy systems – offering an important application for developing, as well as contributing to, understanding the opportunities and challenges of RI in practice. However, while the 12 LLs have been successful in ensuring responsiveness, and promoted a shared responsibility for structuring and co-designing participatory interactions (Baran and Berkowicz 2020), the expectations that the processes initiated would continue to flourish after the end of the planned activities, were not always met. This indicates a lack of sufficient attention paid to anticipation and responsiveness in LLs, which often produce innovations that are not effectively integrated into policy (Lewis 2021). In future research, to successfully ensure that co-created, validated ideas and proven concepts can be implemented in real-life policies, it seems important to enlist early those who will be responsible for applying new policies or solutions (e.g. policymakers, investors) as active participants in the process. Still, this may pose further strains on engagement given the increased heterogeneity of participants (Chilvers 2013; Franz, Tausz, and Thiel 2015). Indeed, in the 12 LLs studied, the ability of researchers to meet the expectations and sustain a continued engagement was challenging.

The analysis equally shows that without adequate reflexive tools embedded in the methodology, LLs risk becoming spaces for innovation that exacerbate pre-existent inequalities (Evans and Karvonen 2014). Reflexivity needs to also consider the issues of time and income, as some LL participants were involved on a volunteer basis and were not effectively compensated for their participation. This prevented a more systematic reflexive process and participatory involvement. Future LL initiatives should therefore ensure a budget for participants (an issue often overlooked in budgeting research projects). Higher reflexiveness in future research can benefit from bringing further attention to the LLs’ responsibility to promote new products, ideas and governance arrangements that are in line with local social, environmental and economic sustainability aspects (Adam and Groves 2011; Bäckstrand 2006; Chilvers 2013). Furthermore, ongoing reflexivity is critical to effectively monitor and anticipate how other social and natural systems, not directly relevant to the LL activities may be either negatively or positively affected by them.

Finally, in the scope of a sustainable energy transition, LLs resulted in advancing and accelerating local processes, effectively supporting the development of new energy communities and prosumer projects. These new projects were sometimes careful to consider other social (e.g. energy poverty) and environmental aspects, that can be addressed through new decentralised renewable energy systems, such as interrelations with water and land use management (i.e. HR1 and PT1). A significant benefit for those participating in the LLs has been an increased energy literacy, which is critical to advance with a sustainable energy transition involving citizens (Sovacool and Blyth 2015). Also, LL participants recognised they gained knowledge of new regulatory frameworks (for prosumers, renewable energy communities and citizens energy communities) (Campos et al. 2020), and on new business models for prosumers, which they have also helped design through participating in business model-focused workshops (Brown, Hall, and Davis 2019). Participants also learned about the application of new technologies, including new approaches to modelling energy communities, such as the one proposed by (Pontes Luz and Amaro e Silva 2021), which resulted from one of the LL experiences.

Conclusion

Living Labs have been found to potentially foster the key dimensions of RI (inclusion, anticipation, responsiveness, reflexivity, and sustainability). Given that all LLs in this study aimed to contribute to more sustainable and prosumer-centred energy systems, the sustainability dimension of RI was more prevalent and self-imposed throughout the diverse experiments.

Nevertheless, this analysis found several limitations, namely that incorporating the RI dimensions into LL approaches requires a conscience awareness of inclusivity, anticipation, responsiveness, and reflexivity as guiding aspects. Failing to do so, risks setting up LLs which reinforce pre-existing structural inequalities and thus help further perpetuate injustices, rather than fostering new solutions that effectively tackle existing problems. Implementing measures that foster from early on a reflexive awareness and a sense of care among LL practitioners is therefore of key importance to ensure RI processes that effectively lead to overcoming injustices and foster inclusivity. Reflexivity is particularly important in the context of the energy transition, where innovations (e.g. electric vehicles, prosumer initiatives) are not equally accessible to all. Also, the development and implementation of new technologies, such as smart neighborhoods and distributed energy systems, needs to be supported by guidelines that target the more vulnerable and disenfranchised citizens, and ensure at the very least that such innovations will not reinforce pre-existing injustices.

In addition, there is a significant and still untapped potential for developing new LL approaches that will contribute to RI in the scope of the energy transition. For instance, a LL can have the sole focus of developing policies and solutions for increasing gender balance within the energy system. LLs can equally offer a forum for addressing those needs that may be at the margins of mainstream energy transition and other decarbonisation policies, such as racism and other intersectional issues. Future energy transition-related LLs could also explicitly focus on addressing climate, environmental and energy justice related problems. LLs may also bring to the foreground restorative justice practices, with a focus on issues such as energy poverty, or addressing transition policies which can have an unequal burden over different communities and social classes.

The energy transition is a moving target and therefore requires flexible and dynamic RI approaches that involve practitioners on the ground, connect citizens, policymakers, and decision-makers at multiple levels of governance, as well as innovators, working across technology, financial, business, and social innovation aspects. Although the focus of the 12 LLs experiences was on RES prosumers and the energy transition, the results of this study are equally pertinent to the application of LLs for RI in other domains of sustainability transitions.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by European Commission: [Grant Number 764056]; Fundação para a Ciência e a Tecnologia: [Grant Number 2020.01663.CEECIND/CP1605/CT0005]; Fundação para a Ciência e a Tecnologia: [Grant Number UIDB/00329/2020]; European Comission, European Union: [Grant Number 764056].

Notes on contributors

Inês Campos

Inês Campos received a doctorate in Climate Change Adaptation and Sustainable Development Policies/Sociology from the University of Lisbon. Her post-doctoral research integrates interdisciplinary social sciences, sustainability and energy research, and data analytics, focusing on the societal aspects of climate change adaptation, including social innovation, and the active role of citizens in the energy transition. She coordinated the H2020 project PROSEU (Mainstreaming prosumers in the energy union). She is presently coordinating the Horizon Europe project INCITE-DEM (Inclusive Citizenship in a World in Transformation: Co-Designing for Democracy). Her main research focus is on delivering new knowledge for a more just and inclusive socio-ecological transformation. She enjoys navigating between quantitative methods and data analytics and participatory approaches to building a broader societal engagement in climate change mitigation and energy transformations. Esther Marín-González is a researcher at the Centre for Ecology, Evolution and Environmental Changes (cE3c) located at the Faculty of Sciences of the University of Lisbon (Ciências ULisboa, Portugal). She received her doctorate in Biology and Plant Biotechnology from the Autonomous University of Barcelona. She also holds a MA in European Studies on Society, Science and Technology (with a specialization in Governance, Innovation, and Sustainability), and a MSc in Scientific, Medical and Environmental Communication. Her current research interests are in the interplay between science, technology and society, science and environmental communication, and knowledge co-production in participatory research practices.