Laboratory settings as built anticipations – prototype scenarios as negotiation arenas between the present and imagined futures

ABSTRACT

As opposed to the recent tendency in Responsible Research and Innovation and in some approaches of Technology Assessment (like Vision Assessment) to reduce the role of socio-technical visions and scenarios to their impact on present debates, our contribution argues that a specific type of future concepts – situational scenarios, and especially their manifestation as prototype scenarios – should be conceptualised as hybrid realities and as negotiation arenas between the present and imagined futures. Based on empirical evidence from the field of ubiquitous computing we apply this concept for analysing three major functions and uses of situational scenarios in the process of technology development: specification, evaluation, and demonstration. We argue that recalibrating the relation between the present and imagined futures is an important aspect of all these functions and uses of situational scenarios, especially when they occur as prototype scenarios.

KEYWORDS:

• Socio-technical futures
• scenarios
• prototypes
• situational scenarios
• prototype scenarios

1. Introduction

The reality as constructed by situational scenarios and especially by their manifestations as prototype scenarios shares striking similarities with the reality of Louis Pasteur’s vaccine against anthrax in his laboratory as described by Bruno Latour. In his seminal article ‘Give Me a laboratory and I Will Raise the world’, Latour (1983) points out how not only the vaccine itself, but also the carefully modified reality within the laboratory was crucial for Pasteur’s approach. In order to make his vaccine work, Pasteur implemented in his laboratory an artificially constructed environment with certain hygienic conditions required for successful immunisation of cattle against anthrax. This ‘reality within the laboratory’, in which it is possible to immunise cattle against anthrax – but only those living under particular laboratory conditions –, is a peculiar hybrid phenomenon. It is a hybrid reality located between fact and fiction, between the present and an imagined future. On the one hand, it is an established reality: in the laboratory the vaccine actually works. On the other hand, this reality within the laboratory refers to a future not yet realised: a future in which the vaccine works not only within the laboratory but also ‘out there’ at the farms where it is needed for practical purposes. This is the kind of hybrid reality between present and imagined future we will refer to in this article.

The focus of our empirical work on analysing technology development in the field of ubiquitous computing (or ambient intelligence, cyber-physical systems) is on how scientists and engineers in their actual work of inventing new technological methods, processes and devices are influenced by imaginations about the future. To address this question we believe it to be useful to focus on imaginations that envisage the future in a more concrete way than visions usually do (cf. Schulz-Schaeffer 2013). We identify a special type of future concepts that are especially influential when it comes to informing concrete conceptual and design decisions of technology developers. We call these situational scenarios. Situational scenarios are images of the future that specify in some detail for imagined typical situations of using an envisaged new technology how the components of these situations – the features of the technology, the users with their interests, preferences and capabilities, other people, objects or structures of relevance for the situation – would (or might) interact (cf. Schulz-Schaeffer and Meister 2015, 166). Situational scenarios are present in the innovation work we studied in different manifestations: as narrative scenarios, that is, as symbolic representations of the respective scenario as written story, video, or comic; as implicit scenarios, existing only as more or less tacit mental images, and thus observable only by their effects on the engineers’ work; and in a form we call prototype scenarios. Prototype scenarios are more or less comprehensive implementations of the components that make up the scenario within more or less realistic laboratory settings.

Our claim is that situational scenarios and in particular their manifestations as prototype scenarios represent hybrid realities that are quite similar to the hybrid reality in Pasteur’s laboratory. As we will show, prototype scenarios possess unique features as negotiation arenas between the fictional reality of an imagination and the empirical reality of the existing world. This is because prototype scenarios participate in both realities simultaneously.

The considerations of this paper build on about 60 interviews with researchers from the field of ubiquitous computing, which we conducted 2013 in the USA, in Japan, and in the EU. Our findings rely on grounded theory-based interpretations of the transcribed interviews combined with content analyses of scientific publications and related research documents. We gathered the material as part of the research project ‘Scenarios as patterns of orientation in technology development and technology assessment’ (funded by the German Research Foundation, DFG).

The remainder of this paper is structured as follows. In Section 2, we put our considerations into the context of two major strands of conceptualising future concepts in Science and Technology Studies (STS), Technology Assessment (TA), and Responsible Research and Innovation (RRI). We present a different conceptualisation that views socio-technical futures as negotiation arenas between the present and imagined futures, using Latour’s reconstruction of the ‘Pasteurization of France’ as an illustration. In Section 3, we define in more detail situational scenarios and prototype scenarios that we identified in our empirical investigations in the field of ubiquitous computing and describe some of their cognitive functions and uses in technology development. In Section 4, we show how prototype scenarios serve as negotiation arenas between the present and imagined futures for three major functions and uses of imagined futures in processes of technology development, namely specification, evaluation and demonstration. In the concluding section, we summarise our empirical findings and conceptual considerations.

2. Two major poles of conceptualising future concepts and a third position in between

Conceptualising imagined future situations as negotiation arenas between the present and possible futures positions them between two major poles of conceptions from STS, TA, and RRI. A large part of work in STS on images of technological futures deals with the governance of new technologies. Here, STS scholars address how the proponents of scientific and technological progress employ images of the future in order to attract attention and support for their purposes. They analyse how images of desirable or dystopic sociotechnical futures affect the beliefs and perceptions of the public, and they look at how today’s political, economic and societal decisions concerning technological change are influenced by such images of the future (see e.g. Borup et al. 2006; Jasanoff and Kim 2009). Most of this kind of research focuses on technological visions, that is, on broad and general ideas about technological futures like ‘carbon-free energy’ or ‘smart cities’. In contrast, concrete pictures of future technology and its societal consequences are of course common to the tradition of TA. Methods of foresight, horizon scanning or scenario planning all try to spell out such a concrete picture of the future (or plausible paths towards such a future), and traditional TA methods aim at securing the adequacy and probability of the futures portrayed. Even for scenario analysis, which focuses on ‘backcasting’ from projected future situations to present decision-making (c.f. Dreborg 1996; Börjeson et al. 2006; Quist and Vergragt 2006), assuring the plausibility of these future situations is at the heart of the approach.

Recent discussions have shifted the emphasis from portraying a probable socio-technical future (or plausible alternative futures) to the present role of any portrait of the future, often combined with a strong critique of ‘traditional TA approaches’. In RRI, which aims at designing innovation processes to be as open as possible regarding contributions and content, this is apparent from the very definition; see the following quote from one of the founding papers: ‘Responsible innovation means taking care of the future through collective stewardship of science and innovation in the present’ (Stilgoe, Owen, and Macnaghten 2013, 1570). And even though ‘anticipation’ is commonly introduced as one of the four dimensions of RRI (besides reflexivity, inclusion and responsiveness), there are strong warnings about any ‘prediction’ and the risk of ‘exacerbating technological determinism’. In short: ‘Any process of anticipation […] faces a tension between prediction, which tends to reify particular futures, and participation, which seeks to open them up’ (Stilgoe, Owen, and Macnaghten 2013, 1571). Imagined futures in RRI are strictly seen as part of present ‘capacity building’ (Stilgoe, Owen, and Macnaghten 2013, 1570), and ‘anticipation’ is but a cipher for rejecting any concrete picture of the future, to keep the image of socio-technical futures as open as possible for present negotiation and inclusion.¹ In Vision Assessment, which mainly focuses on visions of large-scale breakthrough innovations (like nanotechnology or human enhancement) the envisaged socio-technical futures are viewed in a similar way: ‘The main goal (of vision assessment) is to investigate the roles and functions of visionary future communication in the ongoing debate on the future of society and of humankind’ (Grunwald 2009, 147; see also Grin and Grunwald 2000). Thus, analysing and assessing the influence of sociotechnical futures on the present is a concern shared by many of the contemporary approaches in STS, TA, and RRI (cf. e.g. Lösch et al. 2016).

While we agree that claims of accurate forecasts are not only highly questionable but also may have undesirable political and ethical implications, we are dissatisfied with reducing the role of imaginations of socio-technical futures to their possible impact on present debates on the future. In line with our interpretation of Latour’s Pasteurization of France, we instead suggest conceiving of envisaged socio-technical futures as negotiation arenas between the present and imagined futures.

To clarify in which way socio-technical futures – and especially those that are prototypically realised – can serve as negotiation arenas between the present and an imagined future, let us take a closer look at the hybrid reality in Pasteur’s laboratory. Within the laboratory, vaccination against anthrax is a reality. However, outside the laboratory on the farms, where the cattle are still dying from anthrax, it is not; rather it is an imagined future yet to be realised. According to Latour, to become a reality outside of the laboratory, the cattle farms had to be ‘in some crucial respects transformed according to the prescriptions of Pasteur’s laboratory’ (Latour 1983, 151). Latour describes this process of transforming the reality of the laboratory into a societal reality as a negotiation between Pasteurians and representatives of agricultural interests around the problem of finding ‘a compromise that extends Pasteur’s laboratory far enough – so that the vaccination can be repeated and work – but which is still acceptable to the farming representatives’ (Latour 1983, 151). It is a negotiation process in which the existing reality of the farms is changed according to the new network of relations established in Pasteur’s laboratory but in which the new structure as well is adapted to the reality of the farms.

In our interpretation, Latour’s reconstruction of the ‘Pasteurization of France’ (Latour 1988) is about how an imagined future reality is transformed into present reality by means of constructing a reality that lies in between and serves as a negotiation arena for both realities. It is about how the envisaged socio-technical future of defeating the widespread anthrax disease in nineteenth-century agriculture in France is transformed into reality by means of constructing a hybrid reality in the laboratory.

According to Latour, Pasteur had to make three consecutive moves to transform the imagined future of a world in which cattle are immune against anthrax into reality. In the first move, he had to explore the conditions at the farms in relation to his approach. In the second move, he had to bring the world of the farm into his laboratory, to specify both the substance of the vaccine and the conditions under which it could work. The new socio-technical constellation, including a weakened anthrax bacillus serving as the vaccine and cattle living under hygienic conditions, was a basic realisation of the imagined future situation in which immunisation of cattle against anthrax under agricultural conditions would be a reality. Thus, in the laboratory setting the future of agriculture was already there. But initially, this future reality only existed within the narrow constraints of the laboratory. In its initial state, it would not work in the complexities of the real-world agriculture. Like the simulated snowfall in a snow globe it would immediately cease to exist if the containment of the laboratory was removed. So in order to ‘raise the world’, Pasteur had to take a third step: To get the situation – prototypically realised in the laboratory – out of the laboratory and into society, the cattle farms had to some degree be remodelled according to the prescriptions given by the laboratory setting.

In our interpretation of Latour’s considerations, the future situation which is prototypically realised – that is, the prototype scenario – is influential for the course of technology development as it initiates and structures a negotiation process between the present and an imagined future. To allow for specification and evaluation of the new approach, the reality of the farming conditions had to be built in the laboratory setting – the future thus became reality, but under laboratory conditions. And the imagined future situation structured the negotiation process that led to the extension of the laboratory setting into societal reality.

3. Situational scenarios, prototype scenarios, and their cognitive functions and uses in technology development

Scientists and engineers often refer to imaginations of socio-technical futures when developing new technologies, at least if these developments go beyond strictly incremental innovations. These future concepts often are much less general than visions typically are. Rather, they depict in some detail how the envisaged future might look in reality. Future concepts of this kind specify for imagined typical situations of using the new technology, how the components of these situations – the features of the technology, the users with their interests, preferences and capabilities, other people, objects or structures of relevance for the situation – would (or might) interact. Adopting a term from scenario research, we refer to this kind of future concepts as situational scenarios (cf. Steinmüller 2003, 11; van Asselt et al. 2010, 26–27; Schulz-Schaeffer 2013, 30). According to Herman Kahn and Anthony J. Wiener, the pioneers of scenario research, scenarios ‘are hypothetical sequences of events constructed for the purpose of focusing attention to causal processes and decision points’ (Kahn and Wiener 1967, 6). Situational scenarios (in contrast to developmental scenarios or chain scenarios which are about possible future developments) are descriptions of possible future states of affairs.

One main characteristic of situational scenarios is crucial for our argument: Descriptions of future situations of using imagined new technologies necessarily are descriptions of the interactions and relationships between the components included in the scenario. These components include, but are not restricted to the users, the technology, cultural norms, legal restrictions, other people, objects and circumstances relevant for the imagined situation, and so on. If to count as credible and consistent descriptions of possible futures (cf. Wilson 1978; Godet 1986, 135), they have to draw a picture in which the included components fit to one another and act and interact in a believable way, regardless their level of detail. This characteristic of situational scenarios has two interesting consequences: First, because situational scenarios combine features of a future technology with features of an imagined context of application, spelling out technical features necessitates defining requirements for an appropriate situation of use, and vice versa. This is the basic mechanism behind the guiding role of scenarios in the process of technology development (cf. Schulz-Schaeffer and Meister 2015). And second, if some or all of the components of the scenario are prototypically realised in the engineers’ laboratory (just as Pasteur built a prototypical realisation of immunisation against anthrax in his laboratory), then part of the imagined future becomes actual reality. This puts situational scenarios, and especially their manifestation as prototype scenarios, in a position between the present and the imagined future.

Our findings on the role of situational scenarios and prototype scenarios in technology development are derived from empirical research in the field of ubiquitous computing, a field chosen because researchers and engineers here refer regularly to technological future concepts. Not only is there a vision which describes the overall aim and which is generally accepted by the protagonists of this field; the protagonists’ publications also often contain those more specific and elaborated images of the technological future we call situational scenarios.

The vision of ubiquitous computing was presented at first by Mark Weiser. According to Weiser’s vision, computers will become ‘an integral, invisible part of people’s lives’ (Weiser 1991, 66). A myriad of interconnected computing units embedded within the users’ everyday environment will constitute a constant background presence, a ubiquitous informational infrastructure that is intuitively usable based on everyday knowledge. Right from the very beginning, this vision is spelled out by situational scenarios such as the scenario Weiser presents in his foundational paper on ubiquitous computing: ‘Sal awakens; she smells coffee. A few minutes ago her alarm clock, alerted by her restless rolling before waking, had quietly asked, “Coffee?” and she had mumbled, “Yes.” “Yes” and “No” are the only words it knows’. In her office

Sal picks up a tab and ‘waves’ it to her friend Jo in the design group, with whom she has a joint assignment. They are sharing a virtual office for a few weeks. The sharing can take many forms – in this case, the two have given each other access to their location detectors and to each other’s screen contents and location. […] A blank tab on Sal’s desk beeps and displays the word ‘Joe’ on it. She picks it up and gestures with it toward her live board. Joe wants to discuss a document with her; and now it shows up on the wall as she hears Joe’s voice: ‘I’ve been wrestling with this third paragraph all morning; and it still has the wrong tone. Would you mind reading it?’. (Weiser 1991, 74–75)

In the development projects we studied and analysed, we identified many situational scenarios. Our findings strongly support the assumption that situational scenarios provide guidance in development projects. Furthermore, we claim to be able to show how this works. This of course does not mean that technology development is always driven by scenarios – often purely technical questions (extending the ‘state of the art’) or simply grasping opportunities take the lead. Nor does it mean that all scenarios that can be found actually succeed in guiding technology development, even if they are presented prominently. Often these scenarios are fashionable illustrations that are tailored to convince third parties of the very sense of a possible technological application without any influence on the actual developments.² For instance, there are many scenarios that envision ubiquitous computing technologies as solutions for problems of the aging society. Obviously, part of the reason for this is that this problem is high on the agenda of funding agencies and journalists.

From our empirical data, we derived some analytical distinctions in order to grasp the different forms, functions and uses of scenarios (see for details and some illustrative cases Schulz-Schaeffer and Meister 2015). Accordingly, scenarios can

1. occur as a very brief picture of the application of future technology or as a picture that is extensively spelled out;

2. describe a very specific future technology in a very specific domain of application or describe a generic technological solution that aims at solving a particular class of structurally similar problems; for example, identifying unusual behaviours with sensors and appropriate detection algorithms can be depicted as a solution for many societal problems in many domains, ranging from falls of seniors in their home to gathering typical movement patterns of drug dealers or bank robbers to the military domain;

3. start from a detailed description of the envisioned technological components calling for respective application requirements, or vice versa;

4. occur in different manifestations, ranging from implicit pictures that exist only ‘in the head’ of the engineers to forms of explicit representation (text, picture or video) to material realisation in the engineers’ lab;

5. play out different guiding roles for the development process, covering specification, evaluation or demonstration of the future technology and/or its situation of application.

Because the third and the forth of these points are crucial for our argument, we next give a short sketch of these. We then refer to the fifth point to elaborate on why and how prototypically realised scenarios serve as negotiation arenas between the present and imagined futures.

3.1. Technology-oriented vs. application-oriented guidance by scenarios

Credible and consistent situational scenarios of a future technology in its imagined domain of application have to draw a picture in which the included components – the users, the technology, and maybe other components and circumstances of the situation – fit to one another and thus act and interact in a plausible way. This requirement provides the basis for the third point, as sketched above. Thus, a plausible scenario that envisions a new technology with particular features and functionalities also needs to spell out when, where, why and how these features are useful. That is, it needs to describe or to invent contexts of application for the imagined new technology as well as societal needs for which it provides a solution. Thus, the technical promise to some degree defines requirements for an appropriate application problem. We refer to this kind of scenario-induced specification as application-oriented guidance of situational scenarios.

There is a similar requirement for specification induced by scenarios that work the other way around. Here, the scenario starts with a rich description of a problem within a particular domain of application that evokes the need to specify the features and functions of the technology which is imagined as the future solution to this problem. The fall detection scenario, as described and analysed below, provides an example. In this case the description of the imagined context of application defines the requirements for an appropriate technological solution. We refer to this kind of scenario-induced specification as technology-oriented guidance of situational scenarios.³

In our analyses of the guiding role of scenarios we find that in some cases this guiding sometimes only occurs as one single episode in the course of development and sometimes ranges over longer periods of time. In the latter cases, application-oriented guidance can take the lead over several consecutive episodes (as in the example sketched below); however, there are also cases where the development flips from application-oriented to technology-oriented guidance by the scenario and vice versa.

3.2. Manifestations of situational scenarios

According to the fourth point from our list above, we suggest distinguishing among three different manifestations of situational scenarios: narrative scenarios, prototype scenarios, and implicit scenarios.

Narrative scenarios: As in the seminal example of Weiser’s scenario quoted above, situational scenarios often take the form of a narrative providing a more or less elaborated story in the medium of written or spoken language. The plot of these stories is typically that of a user’s beneficial possibilities created by the future technology. However, these narratives can also be told by other forms of media, such as comics, short film tracks, or pictures. One interesting subclass of these representations are diagrams or sketches in which the interplay of the components of the future situation are presented in principle (see Figure 1).

Figure 1. ‘Schematic diagram’ of a conference support system (project publication 1998).

Note: We do not give the full reference of this and of all the following quotes from publications to maintain the anonymity of the individual researchers (with whom we also conducted interviews) and the research projects.

Prototype scenarios: Engineers who develop new technologies are usually eager to construct prototypes of the technology under development as well as the corresponding testbeds in the early stages of their work. They do so in order to test, demonstrate and evaluate their ongoing work and in order to arrive, as early as possible, at a proof of concept by showing that the technological solution works as intended, at least in principle. Every constellation of technological prototypes and corresponding testbeds embodies an underlying situational scenario. It embodies a particular idea about how the technology – as represented by the prototype – and the users, and other relevant components and circumstances – as represented by the testbed – should interact in typical future situations of use. Thus, every constellation of this kind is a prototypically realised scenario or, as we call it, a prototype scenario. Here, again, the technical side of the scenario (the prototype) and its applications side (the testbed) have to be developed in close interrelation to achieve a high degree of plausibility and believability.

An early prototype that actually works is usually a very basic version of the envisioned technology, a version that lacks not only the nice design of marketable technology but also a considerable part of its functionality. Early prototypes usually are manufactured from components already available or hand-crafted for the specific purpose, and their functionality is limited to the most basic features that make the difference to prior technology.

For instance, one of the Japanese research teams we studied developed a sensor network system whose basic new feature was to remind people with dementia to fulfil everyday tasks as intended by them. To this end, motion sensor nodes combined with red and green LED lights were attached to everyday objects such as cups, water kettles or tooth brushes. From the motion information the system concludes whether the everyday activity is conducted as usual or not. If the patient forgets the next step, the green LED at the object to be used blinks (or the red LED at one that should not be used) to remind him or her what to do next. The sensor node module of the early prototype was a flat device about two-inches long and one-inch wide and thus obviously was far too large to be actually employed in everyday use. The first tests were performed with probands without dementia. Nonetheless, these tests were used to prove that the system was able to identify the actual step in an ongoing everyday activity and to predict from learned information about user routines which step should come next.

The example shows not only that the technological components are but a rough draft of a system that would actually work in an everyday environment, but that this is also true for the application side of the prototype scenario. The testbed in which the prototype was tested included people performing everyday routines in their personal way, so that the technological system could prove its ability to learn these personal characteristics and to figure out which step of an everyday activity actually is performed and which one should come next. But this testbed had to deal neither with real dementia patients nor with all the complexities resulting from the fact that the same object may be involved in different everyday activities. Furthermore, the demonstration did not show how the reminding subsystem actually worked and ignored all the problems related to the much too bulky sensor node and many other differences to the imagined real-world future of this technology.

As the example shows, the demonstration of a future technology in its early stages of development usually includes a technological prototype, which already possesses some of the most important features that distinguishes the new technology from the present state of the art. It also includes a corresponding testbed that, on the one hand, allows for demonstrating that these features actually work and, on the other hand, allows for disregarding all the aspects of the technology and of the context of use that are not yet solved or taken into account. Thus, the resulting prototype scenario can comprise components that at the present time are already highly elaborated, detailed, and realistic in some respects and at the same time remain rather vague and crude, or even unrealistic, in other respects.

Prototype scenarios are often used in the engineers’ laboratories in a systematic way over longer periods of time. Additionally, engineers’ construct for demonstration purposes one-shot instantiations of the scenario, polishing up the messy ongoing laboratory work for a test or a demonstration, where the imagined future situation of use is shown to a wider public. The following pictures (Figures 2 and 3) depict this difference between the work in the laboratory and the version of the system for public demonstration:

Figure 2. Laboratory realisation of a scenario. Title: ‘Structure of Environment Imitating Museum Booth’ (project publication 2003).

Figure 3. The basic device for the scenario as developed in the lab (left; project publication 2003) and as used for a public event (right; publication 2009 from the same project).

Implicit scenarios: In not a few of our cases we clearly saw situational scenarios at work that were not explicitly represented as text, film, drawing, etc., but remained ‘in the head’ of the engineers, existing only as mental images and possibly as the subject of verbal communication. We refer to this subclass of scenarios, that are not explicitly symbolically represented, as implicit scenarios. One way to learn about implicit scenarios is via realisations in the engineers’ laboratories, that materially represent the scenario. And vice versa: The existence of prototype scenarios without prior symbolic representations of their underlying stories point to pre-existing implicit scenarios.

4. Scenarios as negotiation arenas between the present and imagined futures

Just as Pasteur’s reconstruction of some elements of French farms in his laboratory, today’s prototype scenarios are situational scenarios whose components are to some degree real entities with real properties. In contrast to a purely narrative scenario, the principal properties of the technical prototype as well as the principle parameters of the context of use are physical properties and their interactions are real-world phenomena. Being physically implemented, a prototype scenario is obviously more than just a mental image, an idea or a story written, spoken or sketched. However, it is a simplified and fragmentary version of the imagined future reality it embodies. This is also true for Pasteur’s prototypical realisation of immunisation against anthrax in his laboratory as compared to an imagined future situation in which the vaccine works on farms where it is really needed. A prototype scenario is not yet the real-world situation it refers to. This constitutes the hybrid reality of prototype scenarios. Simultaneously, they represent an actual reality and the fictional reality of an imagined future. This puts prototype scenarios in a unique position as a negotiation arena between both realities and their respective opportunities and constraints. This can be shown for three major functions and uses of imagined futures in processes of technology development.

4.1. Specification function of prototype scenarios

Constructing prototype scenarios urges engineers to specify elements and relations that may have been rather vague in the preceding narrative or implicit scenarios. To become components of a prototypical realisation of a scenario, the elements of the envisaged new technology and of its future context of application have to be implemented as real entities with real properties. These properties should reflect the real interactions in which these entities are involved according to the scenario. Otherwise, even the simplified laboratory version of the technology would not work as intended and would not provide the envisioned uses. A prototype scenario requires the existence of an initial version of the new technology that actually works (to some degree) – the prototype. It also requires that the context and the typical situations of use envisioned for the new technology actually exist in space and time. This means that not only the technical side has to be specified to considerable detail, but the application side as well. Thus, the realisation in space and time urges engineers to spell out their imaginations more rigidly than narrative scenarios ever could. So, if it is true that situational scenarios provide guidance for scientists and engineers, because they allow spelling out imaginations about the future in some detail, this should apply even more to prototype scenarios.

The following example of technology-oriented guidance provided by a prototype scenario illustrates this point. The most obvious cases of technology-oriented guidance through situational scenarios are those in which engineers conclude from considering a scenario that the new technology would have to possess certain features that so far did not occur to them. The best evidence that it is indeed the involvement with the scenario that leads to these new ideas is provided by cases in which the engineers’ prior assumptions about the technology’s features become obsolete because of their engagement with such a prototypically realised scenario.

This is what happened when computer scientists from a university on the east coast of the US were working on an in-house monitoring system for seniors. The underlying scenario assumes elderly people who live alone and who are still fairly independent. Their home is equipped with a system of networked sensors for detecting movements. The task of the technological system is to detect unusual deviations from the inhabitant’s daily activities, which would indicate that something is wrong with the person. In such a case the system would notify a relative or a caregiver.

At first, the engineers assumed that what constitutes a regular or an irregular behavioural pattern could be derived from the sequential structure of the monitored activities. They assumed the activities to be monitored would consist of given sequences of basic actions that are conducted in a fixed order, such as cooking, which requires getting the ingredients and preparing them before putting them on the stove. Consequently, the technological system would only have to detect the movements associated with these predefined sequences of action (i.e. from refrigerator to sink to stove) in order to be able to determine regular or irregular behavioural patterns.

However, once the researchers began to build a prototype scenario in which they not only implemented a prototype version of the technological system but also specified the major components of the context of application, they learned that most of the activities to be monitored would not have this predefined structure, but that ‘very important information about the monitored person’s habits […] are often difficult to identify based on a-priori or “common-sense” knowledge’ (project publication 2009). From their experiences with the prototype scenario they learned that most of the regularities in the daily movements of elderly people in their home are regularities based on individual habits. As a result, the engineers developed ‘an algorithm for determining if an event occurs persistently within an interval where the interval is periodic but the event is not’ (project publication 2010). Thus, from the prototype scenario the engineers learned that the technological system needs to learn the inhabitants’ individual habits in order to be able to detect unusual deviations from daily activities while they originally had thought this could be done based on a-priori knowledge about sequences of action. As a consequence, the engineers switched the basic technological approach (and the class of algorithm used) from the recognition of predefined patterns of behaviour to the recognition of typical individual behavioural patterns.

In this example, subject to negotiation is the interaction between the movement detection and recognition system, on the one hand, and the patterns of how people perform everyday activities at home, on the other. The former represents a component of the imagined future situation, the latter a component of today’s situations for which the new technology is designed. The prototype scenario proceeds under the assumption that the behavioural patterns of people at home will remain more or less the same in the future. Consequently – to make this interaction work – the component of the imagined future has to adapt to the component of the present situation. This is not necessarily the result of the negotiations between present and future staged by prototype scenarios. There are more than enough examples in our empirical data where components from present situations are assumed to change according to imagined future components. This holds true not only for technological components but for components of the application situation as well.

4.2. Evaluation function of prototype scenarios

In addition to (and often after) the specification of a future technology and its situations of use, engineers try to verify the overall feasibility of an approach or to test the effectiveness of a promising technological solution (a device, an algorithm, etc., or a combination of these). In our sample we have several cases where prototype scenarios are used to serve this purpose. Prototype scenarios are predestined for this task. Because they are physically implemented, and thus more than (more or less rigid) thought experiments, they take place not only in the mental realm but confront the imagined future with the resistances of present real-world components. Resistance can occur from both sides of the implemented scenario: from the technical features of the prototype or from the wishes, requirements and practices of present users in the testbed.

We illustrate the evaluation function of prototype scenarios with the example of the fall detection project mentioned above. In this European project, engineers created the scenario of a technical system that automatically detects different types of falls of seniors in their homes as the basis for a system that automatically alerts relatives or caregivers at a distance. To this end, the engineers implemented in their laboratory a setting in which predefined falling events of people were simulated and videotaped. This representation of the application side of the scenario, then, was used to assess the accuracy of fall detection of several different detection algorithms and several different types of sensors. The core components of the prototype scenario (the video cameras for creating a 3-D-image of potential fall events and potential fall movements) were implemented in the laboratory setting. Within this setting, the probands performed a series of moves from a list containing the predefined fall events as well as no fall events. No fall events were included under the assumption that a major challenge for a future fall detection system would be to avoid ‘false positives’.⁴ Thus the application components of the scenario in this case define very rigidly the requirements for the technological components. As can be seen in Figures 4 and 5, not only does the prototype scenario contain the most crucial technological components, the behaviour of the probands is also strongly prescribed by the testbed.

Figure 4. Realisation of the scenario in the lab: ‘detection of falls of seniors in their homes with 3-D-camera system’ (project homepage 2010).

Figure 5. List of definitions for movements of the probands in the prototype scenario, defining positive and negative falls (project publication 2010).

Over several years, the video sequences generated by means of this prototype scenario were used for a systematic evaluation of possible improvements of the fall detection components. In at least seven subsequent steps the accuracy of fall detection of different detection algorithms and different types of sensors was assessed quantitatively, not at least including the definition of an adequate threshold triggering an alarm. After the evaluation of different approaches for fusion of pictures from different camera angles, different types of detection algorithms were evaluated and judged. Figure 6 provides an example of one single evaluation step in which the scenario-character of the evaluation setting is obvious.

Figure 6. Setting for evaluation of one approach to fall detection (project publication 2010).

As it turned out after several steps, the accuracy of fall detection (and especially the elimination of ‘false positives’) still was not satisfying, the evaluation activities switched to completely different approaches. For example, as tables or chairs were frequently detected as fallen people, the technical installation in the laboratory was complemented with an overhead camera to better identify the shape of a fallen body. And in one of the final evaluation steps of the project, the data from the overhead camera were used to add a completely different kind of information to the overall detection system: the definition of ‘regions of interest’ where seniors are typically moving in their home, an approach that should allow to identify non-normal situations as an additional indicator for the occurrence of a real fall incident.

We see a very systematic evaluation of the interplay of the components of the imagined socio-technical situation in this continuous use of the prototype scenario. It is obvious that all of these evaluation steps are only possible because the real-world properties of the prototype scenario (as a basic laboratory representation of the imagined future reality it points to) allow for systematically assessing real-world resistances against the approaches, algorithms and technical apparatus chosen, as well as the real-world behaviour of these future components in a present laboratory realisation. In a quite straightforward way, the prototype scenario served systematically as a negotiation arena between the present and the future – as a materially implemented ground for evaluating future technology against present (but domesticated) user behaviour where the material realisation in the laboratory not only oriented, but directly drove the subsequent development of the technological components of the scenario.

4.3. Demonstration function of prototype scenarios

As soon as it becomes apparent that a prototype of the new technology actually works, engineers are usually eager to demonstrate their success to others. In our empirical cases we found evidence that prototype scenarios play a crucial role for these demonstrations. As the following example from the US shows, prototype scenarios used for demonstration are convincing if they draw the attention of the audience to the elaborated and realistic parts of the setting and conceal the yet unsolved problems or let them appear as less important.

The example is about a car driver assistant system that uses context information such as the position of the car or the gaze direction of the driver in addition to natural language understanding in order to support the driver for instance in finding restaurants or parking areas. For the purpose of demonstration, the research team had equipped a car with a prototype of the assistant system. In an interview, one of the researchers told us how a colleague, who was sceptical at first, became convinced by the demonstration:

So […] there is a woman there, she’s been involved in this field for quite a while, and she wasn’t very friendly, we couldn’t work out why. And she didn’t believe the system really worked. She’d seen the videos and said: ‘Yeah, I don’t think it works.’ And so we took her out in the car and were driving down and she’d say: ‘Ask something about that.’ And so we asked it and then she said ‘Well, how about over there?’ And this is what came back: ‘On your right is a Subway.’ And she said: ‘No, there’s no Subway down Hope Street.’ And then her partner said: ‘Well, actually, there is, it just opened a week ago.’ And she was: ‘Okay.’ And then I talked to her a little bit later at a conference and she said: ‘I was very surprised that a system like this actually works, because […] about three years ago […] we had mocked up a video of the way of future human-computer-interaction within the vehicle […].’ And there was sort of this dream video that they made, and it looked very close to what we had actually done. And she was surprised that people had actually implemented some of the dreams that they had had. (UC PL 115, 226–240)

The system is convincing because it seems to work in the real-world. In contrast to a narrative representation of a scenario, it is the working technology that makes the difference. Actually, however, the driver assistant system works in the same restricted sense the dementia patient reminding system described earlier does. For instance, it knows only about restaurants within the limits of a few blocks of a downtown district; beyond these few blocks it is completely lost. It only worked within this few blocks because only this area was adequately mapped in the prototype. As our interview partner puts it: ‘We went down and created these semantic maps [of the area], and for the reason of just demonstrating or getting people to think about. Look at the technology today, it is a very small demo, but imagine the future’ (UC PL 115, 194–198). And, as in the case of the reminder system prototype, there are countless other aspects that distinguish the situation of the prototype scenario from a real-world situation of using a driver assistant with the features envisioned by this scenario. Nevertheless, the prototype scenario gives the audience the impression that the most important steps towards realising the new technology have been taken and this is what makes the demonstration convincing. Demonstrations by means of scenarios can thus effect a recalibration of both the present and of the imagined future: the world as it is becomes less taken-for-granted while the future reality of the scenario appears to be less fictional. We claim that this is another kind of negotiation between the present reality and the reality of imagined futures, a negotiation that addresses what people believe to become real vs. what remains fictional.

5. Summary

According to the science fiction writer William Gibson, ‘the future is already here – it’s just not evenly distributed’. This aphorism captures aptly the presence of the future in prototype scenarios. In prototype scenarios, the fictional reality of future technologies and their future uses become transformed into the empirical reality of a technology that actually works within a particular context of application. However, it is an empirical reality which is encapsulated within the containment of the laboratory and which is not yet societal reality outside the laboratory.

This hybrid reality turns situational scenarios, and especially prototype scenarios, into negotiation arenas between the present and imagined futures. Situational scenarios urge engineers to spell out the interactions between the components derived from present situations and the imagined future components. Especially in the case of prototype scenarios, this leads to serious negotiations between present and future components because here all these components are to some degree real entities with real properties.

Conceptualising situational scenarios – and especially their manifestation as prototype scenarios – as negotiation arenas between the present and imagined futures positions them between two poles of recent scholarly debate. In contrast to the more traditional approaches of TA, these scenarios and their uses in technology development do not try to forecast what will or may happen in the future. They are rather about turning imagined futures into present realities. In contrast to the tendency of RRI and many recent approaches of TA to limit the influence of future concepts to their contribution to present debates and decision-making processes only, these scenarios in a way anticipate the future. They anticipate imagined futures by realising them in a nutshell. On the one hand, prototype scenarios work as described only because they systematically root the imagined elements of a future application in a working present realisation of this application, and thus in present technological possibilities and use practices. On the other hand, their cognitive functions as means of orientation require rearranging these elements from the present together with the imagined elements as components of a future that is – prototypically – already here.

With respect to the specification function, we argued that technology-oriented or application-oriented guidance provided by situational scenarios lead to specification processes that can be described as negotiations between components derived both from present situations and the imagined future components. We provided the example of a prototype scenario of an in-house monitoring system, in which some relevant specifications of the behavioural pattern detection and recognition system (a future component) is derived from real-world properties of everyday behaviour at home (a present component).

The fall detection system development was our example for illustrating the evaluation function of prototype scenarios. Here, we could see that the prototype scenario was systematically used to evaluate the interplay of the components of the imagined socio-technical situation. The example strongly suggests that thorough evaluations of this kind require the real-world properties of a prototype scenario in order to systematically assess real-world resistances against the imagined future components as well as the real-world behaviour of these future components. Using prototype scenarios for demonstrating that the future is already here is clearly about negotiating how real or fictional things are. This is what our example of the car driver assistant system shows. The prototype scenario gives the audience the impression that the most important steps towards realising the new technology have already been taken. We argue that prototype scenario demonstrations are efforts to recalibrate perceptions of reality: the world as it is becomes less taken-for-granted while the future reality of the scenario appears to be less fictional.

The functions and uses of situational scenarios, and especially of their manifestation as prototype scenarios described in this paper, heavily rely on the hybrid reality represented by these kind of future concepts – on the mixture of imagined future components with elements of the existing world they consist of. Clearly, prototype scenarios do not imply that the future is entirely ‘already here’. Just as Pasteur’s prototype of the vaccine only worked within the borders of the laboratory and not in the complexities of real-world agriculture, in prototype scenarios the technological prototypes are but rough drafts of the future technologies they represent and the prototypical situations of use are drastically simplified. But in this rough and simplified versions, the components representing the future, may have strong impacts on the development within and beyond the laboratory.

Moreover, each of these three functions of prototype scenarios can be interpreted as recalibrations of the relation between two different versions of reality: the present reality represented by the world as it is usually taken-for-granted, and the imagined reality of possible futures opened up by science and technology, anticipated by visions, spelled out by situational scenarios and made manifest in prototypical realisations. As stated above, in demonstrations this recalibration has the effect of the world as it is becoming less taken-for-granted while the future reality of the scenario appears to be less fictional. Specification and, more systematically, evaluations embed parts of the future situation in the present reality by means of a realisation of components that have to actually work together. Thus, there is a mutual adaptation of aspects of both realities.

Of course, there is no smooth path from the laboratory version of a new socio-technical situation to its implementation and adoption in society. Just as the transformation of the French farms according to the prescriptions of Pasteur’s laboratory, which brought about many struggles and ended in some compromise, the path of further development of an actual prototype scenario is and shall be contested and a discursive issue. But reducing the specific features of prototype scenarios and their actual or potential functions and uses in socio-technical development to an issue of present debates and decision-making processes drastically narrows the possible scope of analysis, and potentially decouples analysis and practical use of scenarios (and of all kinds of future concepts) from the main protagonists (engineers) and the main locations (laboratories) of socio-technical innovation.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes on contributors

Ingo Schulz-Schaeffer is a full professor for sociology of technology and innovation at the Technical University of Berlin. His recent research in the field of science and technology studies includes: the role of future concepts in technology development, practices and technologies of cooperation in transnational projects of software development, crowdfunding as gift exchange, and locative media.

Martin Meister is a research associate at the Technical University of Berlin. His main research interests are the sociology of technology and science and technology studies, and more generally sociological theory. Actual empirical focus is on Ubiquitous Computing and Robotics.

Additional information

Funding

This work was supported by Deutsche Forschungsgemeinschaft [grant number SCHU 1384/4-1].

Notes

1. Likewise, ‘anticipatory governance’ (Guston 2014) is portrayed as ‘opening’ present discourses and societal negotiations.

2. Whether the scenarios found in the empirical material are but fashionable illustrations or in fact influence the course of development cannot be seen by looking at these scenarios only. These questions can only be decided on the basis of a thorough reconstruction of the long-term development activities of the respective project team.

3. The driving force behind the application-oriented and technology-oriented guidance of situational scenarios is quite similar to the one that keeps the ‘promise-requirement spiral’ of van Lente and Rip (1998, 223) in motion.

4. A ‘false positive’ is a false alarm. The engineers had to learn, from extensive talk with professionals from the sector of elderly care, that missing one critical fall event is not the most challenging problem. The real problem is: No one would use a system that alerts rescue or relatives several times every day without a serious cause. So here the engineers learned from talking about present problems and practices about the basic requirements of their future system.