Keywords

4.1 Introduction

Risk and safety issues in (high-risk) industries have to respond to a variety of (long-term) developments and changes, some of which are described as crises—e.g., the climate crisis. Related weather phenomena, such as droughts, heatwaves or floods, are expected to become relevant for maintaining critical infrastructures and related activities, ranging from ensuring healthy working conditions in hot periods to issues of safe operations in times of drought (Bieder and Villena-López 2022, see also Le Coze and Tillement in this volume). In addition, long-term trends such as digitalization or demographic changes are relevant, as well as the need for a fundamental reorientation of high-risk industries toward more sustainable operations. Regardless of whether they are framed as crises or as long-term developments, recent trends have in common that they generate new complexities and are associated with many uncertainties that must be taken into account when dealing with the safe operation of facilities and infrastructures. Against this background it is argued that empirically and conceptually new answers and approaches are needed in order to address new complexities and uncertainties as well as transformation processes that challenge risk and safety management (e.g., Bieder and Villena-López 2022).

In this chapter, I will follow up on this diagnosis and propose a perspective or approach that could be useful from both an analytical and a practical angle, allowing to address issues of (new) uncertainties, unavoidable surprises and non-knowledge: real-world experimentation. Such a perspective is not new in the field, as several authors have applied the experimental metaphor to explain processes and structures in high-risk industries (e.g., Felt 2017; Parotte 2020).

In the next session, I will clarify how the idea of (collective) experimentation is currently debatedFootnote 1 and carve out main characteristics of concepts of experimentation. On this basis, I will then discuss the potential of an experimental perspective for the field of risk and safety studies from (a) an analytical point of view and (b) from a practical point of view.

4.2 Real-World Experiments, Real-World Laboratories, Collective Experimentation

Over the past decade, there has been a remarkable increase in interest in experimental concepts beyond scientific contexts (Weiland et al. 2017). Terms such as living laboratory, social innovation laboratories, transition experiments, urban living laboratories, real-world laboratories, home laboratories or collective experimentation have mushroomed in scientific and policy debates. What these terms have in common is an understanding of experimentation outside controlled spaces of scientific laboratories. This understanding dates back to the early 1920s, when the Chicago School of Sociology based on the work of John Dewey, Jane Adams and others invented the idea of social experimentation as a scientific research strategy aimed not only at knowledge production but also at the direct application of knowledge in order to improve living conditions in urban neighborhoods (Gross 2009). From these ideas evolved two strands of research that determine the current debate.

4.2.1 Collective Experimentation

The idea of collective experimentation has gained attention in the academic field of science and society studies and in the context of technological innovation in recent decades. The starting point is the observation that scientific practices of knowledge production and technology development are not confined to scientific laboratories, but “burden” society with the uncertainties inherent in knowledge production, in particular with the effects of new technologies that can only be seen when the technologies are in use (Felt and Wynne 2007; van de Poel et al. 2017a). Due to unknown and partly irreversible effects on society, several authors characterized the innovation of technologies as well as the management of technologies as social experiments (van de Poel 2016). Already in 1994, Krohn and Weyer described a tendency to extend research processes and associated risks beyond scientific laboratories into the wider society (Krohn and Weyer 1994). While in those early days the debate was dominated by criticism of such a dissolution of boundaries, the focus shifted to understanding these experiments as opportunities for learning and demanding reflection on how they can be designed in a democratic way. In this sense, Latour (2011) emphasized the active role of society in scientific knowledge production and collective experimentation. Society and different groups of actors become active participants, for example, in the form of citizens’ initiatives questioning wind turbines and other energy technologies, collecting data on environmental pollution or seeking research on rare and orphan diseases (e.g., Gramaglia and Babut 2014; Callon et al. 2009).

Authors who take up these ideas do not expect that it is possible to confine experiments and their outcomes to controlled spaces such as laboratories, whether in scientific buildings or defined urban areas, and clearly point to the limited possibilities of controlling the process (e.g., Gross 2010; van de Poel 2016). Society is inevitably exposed to foreseen and unforeseen developments, the positive and negative outcomes that scientific and technological developments have (Weiland et al. 2017; van de Poel et al. 2017a). Against this background, there is a call for greater public participation in technological innovation processes and thus a democratization of technology development. Thus, this strand of research focuses on exploring the relationship between science and society in the process of generating knowledge, which cannot be bounded temporally or spatially.

4.2.2 Urban Living Labs

Furthermore, there are practical, action- and solution-oriented transdisciplinary research approaches. These aim to provide answers to increasingly complex problems and related societal challenges, such as climate change, transitions in energy and transportation systems or demographic changes. They seek to accelerate transformations toward more sustainable societies (Schäpke et al. 2018). The Urban Living Labs Handbook refers to a crisis situation and understands urban living laboratories as test fields in which responses are generated to pressing challenges cities face, such as adapting the built environment to tackle extreme weather events like heatwaves or floods or decarbonizing transport through electronic vehicles and by reducing individual transport. These laboratories are intended as real-world contexts for designing, testing and learning from innovations in real time. Technical innovations are addressed as well as innovations in services, processes and new networks of actors (McCormick and Hartmann 2017; Marvin et al. 2018). Knowledge production and learning happen in intentionally created, spatially and temporally clearly defined spaces, and it is expected that results remain within defined boundaries unless they are intentionally transferred to other contexts. A strong emphasis is placed on the integration of diverse actors from science, policy and society; and societal actors are given an active role in these experimental settings as active co-creators of knowledge (McCormick and Hartmann 2017).

Experimentation beyond scientific contexts shares some understandings that can be considered key characteristics of experimentation beyond the laboratory:

  • Experimentation in and with society is understood as a way or means of addressing complex socio-ecological and socio-technical challenges that are fraught with many uncertainties (non-knowledge is normalcy not an exception).

  • The focus is on knowledge production, application, learning and revision. Failures are taken as opportunities for learning.

  • Experimental settings are based on the acceptance that answers to certain questions may be surprising and unexpected, that answers are tentative, and that new knowledge may change the overall envisioned goal.

  • Structures and processes in place make it possible to change strategies if necessary or to modify certain aspects.

  • The integration of different perspectives and types of knowledge is central for experimentation beyond the scientific laboratory. And thus, the integration of societal actors is emphasized.

Experimental approaches are expected to provide better responses to the complexities and uncertainties of current developments and transformations than traditional approaches and thus to allow for better management of complex and uncertain situations. Experimental approaches are not reckless and do not ignore existing knowledge or established tools, but they put emphasis on what is not known and on dealing with unexpected developments (e.g., Gross and Hoffmann-Riem 2005; van de Poel et al. 2017a; Marvin et al. 2018; Bulkeley et al. 2016).

4.3 Experimentation as an Analytical Lens in Risk and Safety Studies

Current developments in the field of risk and safety (industry and infrastructure) are complex and fraught with many uncertainties. They require the transformation of practices and processes while keeping critical systems functioning and services operational and secure. The diagnosis is that existing tools and established theoretical concepts are not sufficient to describe and address the complex challenges and associated uncertainties. For example, the dominant safety model assumes that zero uncertainty is possible and understands it as a prerequisite for controlling risk. As a result, the dominant safety culture stigmatizes uncertainties, non-knowledge, lack of control and surprise and thus misses opportunities for learning (Bieder and Villena-López 2022).

For several contexts of risky technology, authors have used the experimental metaphor as an analytical lens. Bleicher and Gross (2016) analyzed the use of geothermal energy using experiments at the household level, an endeavor fraught with many knowledge gaps and potential risks. They showed that actors rely on experimental strategies (without naming them as such) when dealing with unexpected developments. The experiments’ boundaries and related questions of (un-)certainty are continuously negotiated and refined by raising new questions, e.g., on observed temperature anomalies in city centers, questioning definitions and drawing further actors in. The authors found evidence that actors develop an attitude of awareness that allows them to broaden their horizon of expectation and be open for unexpected developments and willing to change strategies when needed. Furthermore, they focused on the question of learning and knowledge transfer and identified two modes of decision-making for incorporating locally-generated knowledge into the overall process of energy transformation: in the expert mode, new findings are directly taken into account and adjustments are undertaken in short time; in the administrative mode, decisions are taken based on standardized criteria and guidelines. New knowledge is taken up rather slowly.

Parotte (2020) applied the experimental metaphor as an analytical lens in a case of high-risk industry—the study of the activities of Radioactive Waste Management Organizations (WMO). Due to related uncertainties, non-knowledge and ambiguities, she conceptualized the search for a nuclear waste repository as real-world-experiment. This perspective allowed her to identify two different mindsets of the experimenting organization which both come into play (to different extents) in the search of nuclear waste repositories: an open mindset that integrates elements of surprise and complexity and allows for changes of initial plans and a closed mindset aiming to control the results of action. Parotte showed that for technical and safety aspects all analyzed WMO had a tendency for a closed mindset. For some organizations she revealed how an open mindset made it possible to include moral arguments and broader (non-expert) perspectives into the deliberative process related to the intended program of the waste depository, but also required the modification of plans. In some of the cases she analyzed the concept of reversibility was introduced in plans and concepts, triggering an open attitude.

Felt (2017) analyzed the period that followed the Fukushima accident as a real-world experiment. The experiment as analytical lens allowed her to better understand how the space defined as a laboratory has continuously been redefined according to new knowledge and which diverse actors and different types of learning were involved in these activities that finally aimed to regain control over processes. She showed how the continuous redefinition of space enabled a more fluid handling of the notions of containment and control which are central in the nuclear industry (Felt 2017: 176).

These research projects reveal that taking experimentation as a lens to analyze organizational structures and processes allows the identification of existing elements (mindsets, strategies, routines, notions, etc.) that are favorable for dealing with uncertainties, non-knowledge ambiguities and complexities. The introduction of new notions such as reversibility and the different handling of notions such as containment allowed for alternative action while maintaining control.

The experimental metaphor as an analytical tool could be applied to cases described in this volume. An analysis of the case of Groningen (see Postmes, this volume), for example, could be of interest to understand if and how industry takes induced earthquakes as an opportunity for learning (the denial of the causal link between gas extraction and earthquakes seemed to have prevented knowledge generation); how local knowledge and perspectives are handled by powerful actors (knowledge and experience of local residents, e.g., on psychological stress, was not taken into account by the gas industry and politics); or whether new knowledge is shared with the local population and with policymakers (the gas industry did not communicate in a transparent and proactive manner).

Applied to the case of small modular reactors in Canada (see Iakovleva, Coates and Rayner, this volume), plans for installing the technology could be analyzed by using an experimental perspective. Such an analysis makes it possible to identify if routines are put in place that allow for the creation of new knowledge and for systematic learning processes that include knowledge beyond expert knowledge in industry and policy (e.g., neighbors’ and public’s experiences). In addition, fuel (uranium) production and waste disposal probably should be understood as part of the collective experiment and thus should be included in considerations using an experimental lens.

4.4 Experimental Design to Maintain Safety in Transformation—A Next Step

An analytical experimental perspective makes it possible to identify and understand if the capacity of structures, processes and state-of-the-art approaches in risk assessment and safety management supports dealing with unforeseen developments and ambiguities and to use them in a productive way for learning and knowledge production. While researchers have attributed the experimental metaphor to contexts of high-risk industry for analytical purposes, the author of this chapter is not aware of a case of explicit experimental organization of industrial projects, concepts or processes (similar to urban laboratories).

Several authors have stated that experimental designs do not prevent the occurrence of adverse developments, but by focusing on continuous knowledge production and adaptation to new and changing circumstances, combined with the acceptance that it is impossible to know everything one hundred percent in advance and the acceptance that things (strategies, plans, actions) can fail, they deliver more robust results and allow for management in highly uncertain situations (Gross and Hoffmann-Riem 2005; van de Poel et al. 2017a; Bulkeley et al. 2016). Existing research has shown that an attitude of preparedness and awareness on the part of the actors and institutions involved is a prerequisite for building and using an experimental design (Overdevest et al. 2010). It allows for adjustments, the provision of resources needed for adjustment and adaptation, and thus preserves the ability to act (Parviainen et al. 2021).

Thus, analyses using an experimental lens reveal points that might serve as a starting point for explicit experimental design in the context of high-risk industry—an appropriate legal framework, application of new notions such as reversibility, or a different understanding of stakeholders’ role. It remains however a task for future research projects to test such an explicit experimental organization in the context of high-risk industry.

The case of Stavanger (see Engen and Morsut, this volume) seems to be a case that might be designed in the form of an experimental setting. The problem of regional transformation and economic change related to the transformation of the oil industry is highly complex and uncertain and will affect the inhabitants and institutions in the area. The multiple dimensions of the transformation—economic, social, but also environmental (due to sea level rise)—could be addressed in an experimental design that integrates actors beyond government and industry in collaborative knowledge production early in the process. Non-experts such as local residents or local businesses are important participants because they will have to deal with new technical and organizational structures in their daily lives. Residents could be encouraged to contribute their own ideas and implement them in small projects on a trial basis, e.g., to test possibilities of alternative means and practices of transportation, and local economic actors beyond the oil industry could participate and try out new business fields. An experimental design would allow for such trials by defining the overall structure and rules accordingly, actively taking into account the legal framework (e.g., Parotte 2020).

In the experimental design, the progress of the projects would be regularly monitored and readjusted if necessary. Projects that fail would not be considered negative failures, but their analysis would provide valuable information on the reasons and allow conclusions to be drawn for adjustments. In terms of research and practice on risk and safety issues, particular emphasis should be placed on safety issues, such as the vulnerability of new infrastructure designs. Safety-related issues should be addressed early in the process and kept in focus during the process in order to identify and respond early to unexpected developments and to integrate emerging issues.

4.5 Concluding Remarks

An experimental perspective seems to offer advantages in dealing with uncertainties and unexpected developments in tackling transformative challenges in the area of risk and safety in industry and related to infrastructures. As van de Poel (2016) argued, an experimental approach for implementing new technologies is a possibility to deal with the control dilemma of technology development. Existing research reveals that an experimental lens as analytical perspective is useful to identify favorable structures, processes and mindsets. Such a perspective does not in itself require change, e.g., of organizational structures or structures of projects, but it can provide the basis for such a change and serve as a starting point for the explicit establishment of (collective) experimental designs that facilitates learning to improve safety for the management of high-risk industries and critical infrastructures.

Using the experiment as a design principle in proactive way could be a further step. However, it is demanding because it involves changes of existing processes and structures or setting them up in a new manner. Although there is no general contradiction between a safety orientation and the experimental approach (see Parotte 2020), an experimental design is challenging because it fundamentally questions established beliefs, such as the culture of zero risk and of controlling everything. It takes time to develop a shared understanding of the experimental approach. Actors need to develop an attitude of openness to uncertainty, and institutions need to provide structures and routines that allow for transparency and collective learning, which includes engaging actors also against existing resistance (e.g., economic interests, interests in preserving knowledge gaps).