Introduction
Technological trajectories are entangled with practices, infrastructures, rules, norms and meanings (Watson 2012). In this vein, Keppler (2018) points to the importance of analyzing the various factors influencing the establishment of a technology. Such an endeavor requires transdisciplinary collaboration among experts in various political, societal, economic, and technological developments. Also, potential user groups are increasingly seen as co-creators of technologies and no longer just as their recipient (Sotoudeh and Gudowsky 2018).
User and stakeholder involvement requires suitable formats and methods that can be applied within dynamic technology development processes (Möller et al. 2021). Methods and tools of technology assessment (TA) already provide various options for foresight and scenario analyses (Grunwald 2024; Hahn et al. 2020; Konrad 2021; Stegmaier 2020). In this paper we introduce constellation analysis (CA), a method for the analysis of socio-technical constellations. Unlike expert or parliamentary TA, CA pursues a transdisciplinary approach aligning with participatory (pTA) and constructive (CTA) technology assessment. These approaches aim to reveal assessments and underlying assumptions of different stakeholder groups in order to integrate them into a collective process of reflection and evaluation (Konrad 2021). As with CTA and parts of pTA, CA evaluates prospective technological design and implementation processes instead of technologies already in use (Grobe 2021). By involving stakeholders with different perspectives and knowledge of a technology, CA – as a soft intervention (Konrad 2021, p. 214) – is intended to assess the opportunities and risks of innovations and to enable stakeholders to incorporate such factors as social or ecological implications into strategy development.
Constellation analysis provides methodological guidelines for transdisciplinary collaboration.
CA provides methodological guidelines for transdisciplinary collaboration (Schön et al. 2007). The method is characterized by simple application, including visualization, which enables access for different groups of stakeholders. Unlike most CTA methods, the focus is not on scenarios or narratives (Konrad 2021; Stegmaier 2020), but on the visualization of the constellation network and its corresponding description.
This article introduces CA and critically reflects this procedure by addressing the research question: How do different methodological procedures of conducting CA affect knowledge integration and strategy development? In doing so, we respond to the call for reflection on methods for transdisciplinary collaboration (in TA) (Defila and Di Giulio 2018; Lösch 2017) and focus on a comparative evaluation of how CA was applied. In particular, we compare the methodological choices made in the process of conducting CA and the effects of these choices. We present results from implementation of CA in two technology development projects, one on weeding robots and the other on automated shuttles in local public transport. Both technologies are still under development and their establishment depends on various interlinked factors (on dealing with automated driving see Fleischer and Schippl 2018).
The results show that CA can be used to build a common understanding among project partners and to refine or adapt the planned project activities on this basis. For this purpose, CA should be oriented to the project duration and the project activities. CA can also be applied to map a bigger picture beyond the project activities and duration and to encourage knowledge transfer. This can be supported by choosing a longer time frame and a clear visualization without too many details.
A bridging concept: constellation analysis
In this section, we showcase the use of CA by focusing on relevant aspects of its application in two technology development projects. A constellation addresses a specific issue or question such as ‘which conditions promote the establishment of a specific technology in a certain area?’ (Keppler 2018). The main component of CA is a constellation map. It visualizes relevant actors, natural or technical elements and systems of signs (such as norms, regulations, ideas etc.), as well as their relations (Schäfer and Kröger 2016; Schön et al. 2007). The elements are considered with equal weighting and analyzed with regard to their relevance for the constellation.
Schön et al. (2007) suggest various notations for the visualization of the relationships. Namely, a line or a circle connects two or more elements with each other, symbolizing their relatedness. A conflict arrow can be drawn to show challenges or a question mark can be inserted for unclear relations or unclear outcomes. The constellation map is complemented by a descriptive text, which contains more detailed information and sources. Both the constellation map and the accompanying text are a result of inter- and/or transdisciplinary collaboration (Schäfer and Kröger 2016) in which persons with different backgrounds, perspectives and expertise identify relevant actors and elements, analyzing and visualizing their relations. Importantly, a moderator guides this process by involving the participants, responding to feedback and building or rebuilding the constellation map. Good moderation is key to involving everyone and facilitating the visualization. Although participants often cannot all be involved on an equal footing due to a lack of time, Schön et al. (2007) emphasize that all relevant perspectives should still be considered.
When implementing CA for applied strategy development, a status quo constellation mapping the current relations, challenges or gaps (e.g., for the establishment of a technology) and a target constellation can be helpful. The latter visualizes a future constellation when, for example, the technology is regularly used by a broad group of people and the intended impacts have been achieved.
Constellation analysis on robotic weed control in northeast Germany
The Uckerbots project
Uckerbots is a transdisciplinary research project to develop and test the prototype of a weeding robot for organic sugar beet cultivation in northeast Germany. The project team consisted of technology developers, agronomists, microelectronic experts and social scientists, who also collaborated with interested farmers. CA was conducted to identify drivers and barriers for the establishment of regional organic sugar beet production in general and weeding robots in particular.
CA was applied in three steps:
(1) Preparation in a core team: The CA was prepared and organized by the social scientists in the project, who are co-authors of this article (Baatz and Schäfer). First, the main question was formulated: ‘What is the status quo of organic sugar beet cultivation in northeast Germany and what needs to change in order to establish regional cultivation and production of organic sugar?’ We decided to map a status quo constellation and a target constellation in which potential challenges have been resolved. We prepared a first version of the constellations on the basis of project documents and publications. An external expert for CA was consulted to enhance the graphical representation of the constellations.
(2) Discussion with project partners: In a second step, we presented the draft to our project partners using a PowerPoint presentation in which the constellations were gradually constructed from the center outwards. We asked them whether any elements or actors were missing in the constellation, and whether they thought the relationships were correctly represented. In this second step, it was decided that the target constellation would take a medium-term perspective, looking five to seven years into the future. This time frame was chosen by the project partners as it should be possible to (partially) implement the identified strategies within five to seven years.
(3) Discussion with farmers and experts: In a third step, we approached five farmers, an expert for regional value chains, and an agricultural advisor with expertise in sugar beet cultivation. The farmers were all located in northeast Germany and interested in cultivating organic sugar beets. The project context was the basis for the recruitment of participants. Similar to the second step, we presented the constellation to participants individually as a part of scheduled interviews and asked for feedback. Insights from the first interviews were incorporated into the later interviews, and the constellations were continuously adjusted after each interview. The CA expert also supported this step.
Results of the constellation analysis
Fig. 1: Status quo analysis in Uckerbots. Green = Natural Elements; Red = Ideas and Concepts; Blue = Technical Elements; Orange = Actors. Source: authors’ own compilation
Constellation analysis on automated shuttles in public transport in northwest Berlin
The NoWeL4 project
NoWeL4 is a transdisciplinary research project focused on developing and integrating autonomous, on-demand shuttles into Berlin’s public transport system to promote sustainable and accessible mobility. The shuttles will primarily operate in the Northwest Development Corridor, a 25-square-kilometer area planned as a low-car zone. The project aims to make the area less dependent on private motorized vehicles from the outset. Researchers, public transit authorities and legal experts are collaborating to address the technical, operational, regulatory and social aspects of integrating a planned fleet of automated, SAE Level 4 [1] shuttles that will offer flexible, intermodal mobility for residents.
CA was conducted in an early phase of the project to structure the key issues and to integrate stakeholder knowledge. The first analysis examined the current status quo, examining the core question: ‘Who or what plays an important role in the successful integration of autonomous shuttles into public transport in the Northwest Development Corridor?’ The second constellation focused on the target state at the end of the project and the possibilities for expanding the results achieved by that point. It explored the question ‘How can autonomous shuttles be successfully integrated into public transport beyond the Northwest Development Corridor?’ The two constellations were mapped separately, one after the other, each in two steps:
(1) Preparation in a core team: Two co-authors of this article (Lange and Linke-Wittich) prepared the constellations together with the interdisciplinary core team of researchers in the ZTG subproject. Documents and relevant academic literature were utilized for this project. The methodological preparation was supported by an external expert. During this process, a preliminary map of each constellation was drafted. This initial map served as the basis for the stakeholder workshop conducted in the next step.
(2) Workshop with project partners and external stakeholders: In the second step of each mapping exercise, a workshop was held with project partners and external stakeholders. Participants included researchers, stakeholders from a public transportation company, public administration, technical experts, legal experts and representatives for people with disabilities. Altogether this was a group of ten people. The workshop was moderated by the external expert. Building on the prepared map, workshop participants were able to engage in in-depth discussions and refine the respective constellations. The project partners decided that the target constellation would relate to the end of the project in 2027.
Results of the constellation analysis
Fig. 2: Extract from the status quo analysis in NoWeL4. Green = Natural Elements; Red = Ideas and Concepts; Blue = Technical Elements; Orange = Actors; Yellow = Legal Elements. Source: authors’ own compilation
Comparison of the constellation analyses
Our comparison of the CAs in the two technology development projects focuses on how the steps of the CA contribute to strategy development and knowledge integration. In both projects, CA proved to be very helpful in structuring the challenges of technology establishment as a basis for developing strategies. It promoted a common understanding of current barriers and ways of tackling them. The consideration of different influencing factors, which can be of varying importance in different technology developments (certain actor constellations, natural factors, regulations), was especially beneficial. The CA helped to structure and visualize the exchange with project partners. Participants in workshops can see directly how their perspective is integrated. The accompanying text was important for a more nuanced description.
The constellation analysis helped to structure and visualize the exchange with project partners.
The procedures for conducting CA in the Uckerbots and NoWeL4 project were similar. Both relied on the preparation of a CA that was presented to project partners and external stakeholders in a gradual construction process. This allowed them to familiarize themselves with the content and the visualization, and to enter directly into a discussion. Furthermore, both CAs consisted of a status quo constellation and a target constellation. Both the researchers from the core team and the project partners found this structure helpful for strategy development.
However, the specific functions that the CA fulfilled in the two projects differed. The Uckerbots project partners had already worked together on a previous project with a similar focus and therefore had a common understanding of the challenges to be addressed in the project. The purpose of CA was to compile these challenges, to visualize the bigger picture (in the sense of looking at the entire value chain) and to make the findings available to potential new project partners. The latter was done in the context of business model development. In the NoWeL4 project, CA served to create a mutual understanding between the project partners regarding the complex challenges of implementing Level 4 shuttles and to refine the strategies within the project context. Some of these objectives only became clear during the mapping process. The different purposes resulted in three key differences in the chosen procedures, which are reflected in the following in terms of their effect on knowledge integration and strategy development:
(1) The chosen time frame for the target constellation: In the Uckerbots project, the target constellation was geared towards a medium-term perspective (5‑7 years in the future) that extends beyond the duration of the project. This allowed the inclusion of project activities such as the design of the weeding robots but also potential future developments such as the use of the robots in crops other than sugar beet or in conventional agriculture. Strategies that are not (or cannot be) addressed within the project were also included. In the NoWeL4 project, the target constellation was more short- to medium-term, oriented towards the project duration until 2027. This strongly linked the constellation with the project developments and ensured that the map is of essential use to the project partners. However, the short time frame also meant that aspects raised by participants that cannot be implemented within the time frame could only be incorporated into the constellation to a limited extent. This led to unequal consideration of project partners and external participants. Some of the external participants, who for instance raised issues of accessibility, had the impression that their perspectives were not sufficiently included. In addition, certain aspects were not addressed at all due to the orientation to the project duration. This can harm knowledge integration. To incorporate the medium-term strategies, the NoWeL4 project partners decided to additionally map a third constellation that covers a time frame at a later point in the project. This was also an opportunity to adjust the short-term target constellation, e.g. by incorporating delays in technological development.
(2) The discussion in one-to-one interviews vs. in workshops: While the participants in the Uckerbots CA were interviewed separately, workshops with all participants were organized in NoWeL4. Both procedures have advantages. Scheduling separate interviews can be easier and more accessible in some fields (e.g. for contacting farmers) and enable collection of additional data for the research project. The workshops conducted in NoWeL4 generated discussions between participants with mutual reference that were not possible in interviews. A workaround used in the interviews was therefore to include insights from earlier interviews in the ones conducted later. However, this still does not amount to an in-depth discussion. The completed constellation was also presented to the Uckerbot project partners once again and they were able to draw new insights from it, e.g. the technology developers learned something about the technical support services required by farmers. The workshops in NoWeL4 proved to be productive for strategy development and knowledge integration because new ideas were generated in co-presence. They also helped to make divergent points of view mutually visible. Project partners provided feedback indicating that the collaborative workshops improved their understanding of how specific project tasks and perspectives of stakeholders are interrelated. This made it clear that an operation with open access for passengers would hardly be feasible within the scope of the project. A corresponding strategic decision was made to focus more strongly on operational and organizational aspects, such as the design of infrastructure aspects, role concepts and technical supervision of the shuttles.
Constellation analysis revealed critical dependencies between individual approval steps and the roles of different responsible authorities.
(3) Depth of detail of the visual representation: While the CA on sugar beet cultivation focused on the most important elements, the CA on autonomous shuttles contained many details and depicted more elements and relationships. In the Uckerbots project, the visualization served as an introduction to the topic. It was shown to potential new project partners in a workshop to find common ground. The workshop was organized with the goal of involving persons interested in developing a business model for the weeding robots. They reported that this served as an effective introduction to the constellation of organic sugar beet cultivation in the Uckermark region. In this way, CA not only served to promote knowledge integration but also to encourage knowledge transfer. The NoWeL4 project used the graphic as the basis for a detailed project plan and for further specialized discussion of the visualized differentiated knowledge. This approach offered a clearer understanding of task interdependencies, stakeholder involvement, potential conflicts, and areas requiring coordination. As a result, it was possible to further refine the project plan. For instance, it revealed critical dependencies between individual approval steps and the roles of different responsible authorities. The discussions regarding the project tasks also clarified the state of technology development for all project partners, which was essential for coordinating project time lines and resources.
Discussion and conclusion
This article deals with the use of CA to assess socio-technical constellations in two technology development projects. It reflects the methodological potential for strategy development and knowledge integration. For the purpose of strategy development, mapping the status quo and a target constellation proved to be helpful in both projects. Based on the comparison of the methodological procedures, we furthermore recommend carefully reflecting on the time frame for the target constellation, the mode of stakeholder inclusion (interviews vs. workshops) and the depth of detail of the mapped figure. Short-term time frames are well suited to enhance the understanding of the project activities and their implementation and, if necessary, to adapt them to the latest developments (e.g. to the current state of technological development). Workshops moderated by an experienced facilitator are a good format for project teams at the beginning of a project who are aiming to develop a common understanding as a basis for further collaboration.
Medium-term time frames allow inclusion of strategies beyond the project duration. During the joint mapping process, the participants can form a common picture of the role of their project activities within this broader range of strategies. A middle ground can be to map three constellations: The current status, the project goals and a target beyond the project. The question of the time frame underlies every foresight process and should be carefully considered (Stegmaier 2020). For the CA of technologies with a low readiness level and therefore high level of uncertainty (such as the examples in this article), an iterative adaptation of the constellation to changed socio-technical conditions can furthermore promote strategy development.
CA offers the advantage that the persons to be consulted can be involved in the mapping process from the beginning (as argued by Lösch 2017), and afterwards by presenting them the completed maps (knowledge transfer). This can be a basis for deepening further aspects, for example by mapping a sub-constellation. A clear, not too detailed, constellation map facilitates knowledge transfer.
