Citizen visions of drone uses and impacts in 2057 far-future insights for policy decision-makers

Aerial drones for civil purposes will expectedly become ubiquitous in society and potentially overfly everyone and everything. This article aims to inform policy decision-makers at governmental and corporate level about how citizens envision drones will impact daily life and society in the future. Such knowledge is critical to ensure the development of sustainable drone solutions. This study includes stories written by 135 citizens in Denmark, from drone novices to experts, describing how they imagine lifeworlds that include drones in 2057. The stories both confirm and extend existing drone use cases. However, we also identify entirely new applications that would require technological combinations and advances, such as drones with motor or cognitive skills that can replace, complement or assist humans by transforming into any desired form, making instant holograms or offering psychological counselling. Thus, the stories provide nuanced perspectives on the potential of implementing drones and showcase the often opposing nature of the envisioned impacts. Overall, our foresight study highlights the relevance of anticipatory governance involving citizens as well as the need for policy decision-makers to deal with paradoxes when setting frameworks for what should (not) be allowed if the use of drones in the future is to benefit all stakeholders.


Introduction
Understanding the potential future (mis)uses and impacts of emerging technologies is of strategic importance to those making decisions at different levels (Chamuah and Singh, 2021;Porter et al., 2004).This is especially so for emerging technologies that are deemed as potential shapers of society (Brey, 2018;Gruetzemacher et al., 2021;Ulnicane et al., 2021).Examples hereof are recent advances of artificial intelligence and unmanned aerial drones that, due to their pervasive nature, might impact every citizen's life (Hwang et al., 2021a;Lidynia et al., 2016;Røpke, 2001) and significantly change the status quo (Abdel-Basset et al., 2021;Bridgelall and Stubbing, 2021;Drexler, 2019;Gruetzemacher et al., 2021;Taeihagh, 2021).These technologies can supplement or fully take over activities that are normally performed by humans (Jacobstein, 2013), and they can be used for both good and bad purposes (Bajde et al., 2022;Buckley et al., 2017;Gersher, 2014;Spiekermann, 2023;Taeihagh, 2021).Their deployment should therefore be based on responsible considerations of the changes they might entail (e. g.Boucher, 2015;Von Schomberg, 2013).The even larger disruptive potential of combinations of technologies further underpins this need (Bajde et al., 2022).This is evident in recent developments of airborne drones that include artificial intelligence which enables them to perform autonomously a range of new tasks and provide many new types of services (Hwang et al., 2021b;Lakshman and Ebenezer, 2021;Spencer Jr. et al., 2019).
Over the last decade, the use of drones for civil purposes has started to take off (Floreano and Wood, 2015;Hwang et al., 2021a;Johnson et al., 2021;Nelson and Gorichanaz, 2019) in the contexts of, e.g.agriculture, healthcare provision, infrastructure inspection, search and rescue operations (Ayamga et al., 2021;Johnson et al., 2021;Sodero and Rackham, 2020;Tan et al., 2021), and in some parts of the world also delivery of goods (Hwang et al., 2021b).However, they can still be classified as emerging according to the characteristics defined by Rotolo et al. (2015): The drone industry is only relatively fast-growing but characterised by a certain degree of coherence persisting over time (e.g.Andersen et al., 2020;Wang et al., 2023); the drones are regarded as technology platforms enabling a variety of also radically new missions in the civil context (e.g.Hwang et al., 2021a;Li and Liu, 2019;Merkert and Bushell, 2020); but like all emerging technologies they are in "the process of coming into being" and becoming important (Rotolo et al., 2015(Rotolo et al., : 1828) ) and therefore shrouded in uncertainty and ambiguity (e.g.Köhler and Som, 2014); and, thus, expected to realise their prominent impact somewhere in the future (e.g.SESAR Joint Undertaking, 2017).
The future of drones is of huge interest to a range of stakeholders, and forecasts, typically made by extrapolating today's contours, show huge economic expectations and a multitude of potential fields of application (Federal Aviation Administration, 2023;SESAR Joint Undertaking, 2017).Such forecasts may include long-term extrapolations but mainly represent educated guesses of what will happen in the short term less than ten years away, and they are updated regularly for the same reason.However, implementing drones full-scale in society in the coming years will expectedly entail changes that will have radical implications >10-15 years ahead, which is the typical time horizon for emerging technologies to gain social relevance (Rotolo et al., 2015;Stahl, 2011).This means that not only expectations of the soon-to-come future, but also far-future perspectivesi.e.>20 years ahead (Bauer, 2018) should be explored, as these can reveal imagined long-term implications of further implementing drones in society (Kwon et al., 2017).Thus, there is a need not only for forecasts describing the path of emergence in a foreseeable future but also for foresight identifying potential far-future implications of implementing drones.Such indications can serve to inspire actions today and help set the direction towards a desirable future with the technology (Porter et al., 2004).This is in the interest of policy decision-makers residing at both governmental and corporate level (Brick et al., 2018), as they can gain a foundation for making substantiated and responsible decisions about the trajectory of drone development (Porter et al., 2004;Chamuah and Singh, 2021)while the management thereof is still possible (Guston, 2008: vi).
The magnitude of changes that drones expectedly will entail and the complexity of the preparation needed call for qualitative studies incorporating multiple perspectives to adequately inform policy-decision makers (Raghunatha et al., 2023).To ensure a sustainable deployment of drones, there is a need for participatory innovation models and the development of balanced regulatory frameworks incorporating the viewpoints of diverse types of actors and considering potential both positive and negative impacts of implementing drones in society (ibid.).This is in line with recent calls for closing the gap in policy decisionmaking caused by "a discrepancy between increasingly long-term consequences of decisions and a myopic-reactive orientation of political institutions and actors" (Bauer, 2018: 36) and seeking to actively shape any technological transformation to fit human and social values in a sustainable way (e.g.Brey, 2018;Spiekermann, 2023).Here, the civil society should be regarded as one of several co-shapers of the world as well as a potential contributor to policy decision-making (Lehoux et al., 2020).And civil society is important to factor in, as drones willwhen they become widespreadpotentially overfly and to some extent affect all citizens.Thus, it is important how citizens envision the future with drones (Boucher, 2014(Boucher, , 2016;;Gersher, 2014;Lindgren et al., 2021;Schweitzer et al., 2019;Wolf et al., 2022).However, their take on the future is too seldom investigated (Boucher, 2016) at least in a European context (Boucher, 2014).
In response to the above-mentioned calls, we have collected farfuture visions of drones among diverse citizens to be able to investigate: What insights do citizens' far-future visions of drone uses and impacts offer to policy decision-makers that can be used for emerging technology governance?By focusing on gaining such insights from a broad range of stakeholders, we tap into the literature stream on anticipatory governance, which by Fuller (2009: 209) is defined as "a strategy to facilitate the acceptance of new technosciences by inviting people to voice their hopes and concerns […] before the innovations are actually implemented".Overall, we contribute insights to those interested in visions of a multiplicity of drone uses as well as potential positive and negative impacts of increasingly implementing these unmanned aerial vehicles in civil society.Furthermore, we have identified 10 overall drone use case categories that can serve to qualify debate about the extent and type of drones to aim for in the future, and we make propositions for what policy decision-makers should consider regarding governance of the technology.We focus on drones as an example of an emerging technology to identify the types and uses of insights that can be gained from citizens' far-future visions.
The remainder of this paper is organised as follows: First, to determine what is important to factor into policy decision-making regarding emerging technology, we present existing knowledge of governance actors and instruments.Next, we review academic literature concerning the future uses and impacts of autonomous drones.Then, we explain the methods used in our empirical study, which was conducted in Denmark; we present the findings; and we compare them to the findings from academic and other accounts.Finally, we discuss the findings, including what the insights generated by our study can offer to policy decisionmakers and make them ponder.

Governance of emerging technology
The rise of new technologiesespecially those expected to become all-pervasive in societyputs demand on policy decision-makers to take the necessary precautions for ensuring a sustainable future (Kaebnick et al., 2016).Accordingly, at governmental level, policymakers must decide on the need for and drafting of laws, regulations, or softer interventions (Arnautu and Dagenais, 2021), which includes choosing the appropriate policy instruments (Howlett and Ramesh, 1993).However, research has for some time pointed to the need for the governing of emerging technology to change (McGrail, 2012).Currently, it tends to be rather reactive and bureaucratic (Bauer, 2018) and neither matches the pace and scale of technological innovation and change nor utilises the informational richness embedded in the perspectives of a multitude of non-governmental actors (Callaghan, 2018;Sarewitz, 2011).A shift is therefore suggested, away from the technocratic mode towards an "anticipatory interventionist paradigm" (McGrail, 2012: 21) considering the contributions of a range of stakeholders (Brick et al., 2018;Lehoux et al., 2020).

Policy decision-makers at different levels
Governmental actors are ultimately the regulators of society and businesses (Bell and Hindmoor, 2009), but those devising and deciding on policy should be seen as a broader entity including also "corporate board members deciding on corporate strategy options" (Brick et al., 2018: 2).Thus, when we use the term 'policy decision-makers', it encompasses actors at both governmental and corporate level.In fact, policy decision-making can also take place at industry level, for example when an industry association decides how businesses in the industry must self-regulate (Steurer, 2013; see, for example, also literature on Big Tech).However, to keep it simple, we delimit non-governmental policy decision-makers to those residing at the corporate level.
Policy regarding emerging technologies is relevant to consider at both governmental and corporate level.Governmental actors need to decide on public policy with the aim to avoid potentially negative impacts of implementing emerging technology (Boucher, 2015).The focus is overall the same at the corporate level, but here decision-makers must ensure that their new technological solutionswhether they develop or acquire themare designed, offered and implemented in a way that mitigates potential negative impacts on society (Spiekermann, 2023).

Instruments of governance
At both governmental and corporate level, in a search for both effectiveness and common acceptance (Capano and Lippi, 2017), decision-making processes involve the definition of goals and appropriate tools of governance (Howlett and Ramesh, 1993).
With the purpose of inducing (or avoiding) change, policy decision-makers at the governmental level can choose certain instruments (Borrás and Edquist, 2013) that have received much attention due to their both theoretical and empirical relevance (Capano and Howlett, 2020).A range of scholars have put forward instrument typologies (see the metareview by Acciai and Capano, 2020), starting with Lowi (1972) and his four policy types (Nicholson, 2002).However, the typology by Vedung (1998) is probably the most widely used (Borrás and Edquist, 2013;Pachego-Vega, 2020) and encompasses many other scholars' suggestions.He distinguishes between economic, regulative and informational instruments and uses the easy to remember phrase 'carrots, sticks, and sermons' to signal how the three instrument classes differ regarding extent of coercion.Overall, governments can seek to change (or preserve) the behaviour of organisations or individuals through financial incentives, by defining what must be complied with to avoid some degree of punishment, or by inspiring through information campaigns and such.Thus, every instrument produces its own effect and "constitutes a condensed form of knowledge about social control and ways of exercising it" (Lascoumes and Le Galès, 2007: 3).At the corporate level, the focus on corporate social responsibility (CSR) -or more recently the, primarily, investor focus on environmental, social, and governance (ESG) aspectshas for decades signalled that businesses voluntarily look beyond short-term profit-making and accept to share public responsibilities (Steurer, 2010(Steurer, , 2013;;Torres et al., 2023).Businesses might self-regulate their operations and innovative initiatives, meaning that they specify the rules and instruments for and monitor their own conductperhaps based on a triple-bottom line management of environmental, social and economic aspects (Albreda et al., 2007;Steurer, 2013).However, CSR/ESG instruments have increasingly been developed by third parties into frameworks, standards, etc., and compliance is oftentimes monitored by independent bodies (Rasche, 2009).In addition, according to Steurer (2013), for the sake of civil society governments might choose to influence the behaviour and actions of businesses by adopting 'soft' regulation that leans on Vedung's (1998) 'sticks, carrots, and sermons' with economic instruments (e.g.subsidies or public procurement), legal instruments (laws without sanctions) or informational instruments (e.g.studies, campaigns or websites), but also hybrid instruments (e.g.public voluntary programmes).Thus, businesses may focus on public affairs and seek to comply with what is in the public interest, and they are perhaps spurred to do so by philanthropic reasons, but certainly due to expectations from both governments and civil society (Steurer, 2013).
Taken together, policy decision-makers must choose the most suitable instruments, decide the customisation of them for the context they are meant to operate in, and perhaps consider a mix of instruments (Borrás and Edquist, 2013).But first, they must understand what is important to factor into their policy decision-making.

From government to (anticipatory) governance
A range of non-governmental actors are increasingly regarded as important for achieving the goals of public policy (Steurer, 2013;Van Lente, 2010).According to Bell and Hindmoor (2009), governmental actors have not lost their capacity to govern, but by changing the focus from government to governance, they seek to make better decisions by including the contributions of non-governmental actors during the policy decision-making process.
As highlighted earlier, businesses can play a key role and be regarded as decision-makers in their own arena.However, also other nongovernmental actors are concerned with and anticipate potential impacts of emerging technology.This broadened view on relevant actors is included in the concept of 'anticipatory governance', where governance should be regarded not only as a governmental focus but as activities that are distributed across a variety of actors with relevant insights (Guston, 2014).The set of actors should include, for example, also scientists and engineers, whose discoveries, decisions and designs may heavily impact the future.Furthermore, there is a need for factoring in the input by non-experts who can also affectand therefore potentially act as co-shapers ofthe world (Lehoux et al., 2009(Lehoux et al., , 2020)).Thus, regardless of whether citizens of any kind can inspire or impact policy decision-making processes individually or through, e.g.civil society organisations, they should be regarded as relevant actors in the policy decision-making process.When seeking to decide on the trajectory of emerging technologies, the perspectives of both governmental and diverse non-governmental actors should therefore be taken into consideration.It may be a demanding, bargain-driven process, but it can maximise consensus in society on what could be a sustainable future with emerging opportunities (Capano and Lippi, 2017), such as drones.

Accounts on the future of drones in academic literature
Like other types of mobile robots, aerial drones are expected to be capable of operating fully autonomously sometime in the future, meaning they will be ablewithout a human in the loopto "determine the actions to be taken to perform a task" (Rubio et al., 2019: 1).With the increased inclusion of artificial intelligence systems that can gather and use data to make decisions, full autonomy might be possible as well as, e.g.cooperative work, friendly human-robot interfaces, and emotion expression and perception (ibid.).Thus, capabilities are expected to increase, meaning that drones might be able to also take on work and social roles (Formosa, 2021).
Although drones are quite affordable and capable of performing many types of missions (Jacobstein, 2013), and although they are used today in contexts "such as agriculture, construction, surveillance, filmmaking, healthcare, energy, environment, public safety and security" (European Commission, 2022: 3-4), the drone service industry in most countries does not show fast growth rates yet (e.g.Frederiksen Harbo et al., 2019).A main reason for this is that significant regulatory constraints meant to safeguard citizens from potential risks are maintained in most developed countries, in e.g.Europe and Northern America, until the necessary technological improvements for ensuring fail-safe flights are in place (Andersen et al., 2020;Floreano and Wood, 2015).
Thus, drones are not commonplace in many societies yet, meaning that there is scarce experience with these unmanned aerial vehicles in public spaces (Cawthorne and Wynsberghe, 2020).However, as drones are regarded as a disruptive technology that will potentially affect everyone (Abdel-Basset et al., 2021;Ilic et al., 2022), the public perception of having drones in the air is continually investigated, e.g. by means of experiments (e.g.Bajde et al., 2017;Cawthorne and Frederiksen, 2020).Such experiments help shed light on what it feels like to be overflown by drones, whereas other studies provide insight into their potential uses and associated impacts.
In this section, we present the available accounts on the future uses and impacts of drones based on a systematic review of scholarly literature.

Literature review approach
To gain an overview of the academic literature on drone futures, we conducted a systematic literature review that followed the methodological advice in earlier research (Snyder, 2019) and implemented the standards for a literature review in the business and management research area set by Tranfield et al. (2003).Three researchers (the authors) constituted a review panel and used a methodological protocol for defining the search strategy in terms of keywords, databases, the perspective of the analysis and in-and exclusion criteria.All decisions were recorded in a review protocol (Tranfield et al., 2003: 215).For the search of relevant scientific articles, we selected the two databases that guarantee the broadest possible coverage of leading social science and business and management scholarly journals to achieve a maximum variation sample: the Clarivate Web of Science and Scopus.Both are known as comprehensive scientific databases (Guz and Rushchitsky, 2009;Goduscheit and Knudsen, 2013); yet most researchers use Scopus in combination with Web of Science due to differences between the databases in journal coverage and timing of article updates (Kivimaa et al., 2019;Wolf and Bernhart, 2022).
We selected search words used in the business and social science literature to label the emerging technology platform, i.e. 'drone*', 'UAS' (unmanned aerial system), and 'UAV' (unmanned aerial vehicle) and combined them with words that relate to our research focus, i.e. 'future' and 'application' or 'use' (see Appendix A).We then used Boolean operators to create meaningful search strings.We tested the keywords in the Scopus and Web of Science databases.As inclusion criteria, we decided to limit our search to publications with titles related to our research topic and describing recent or future drone applications in the fields of Social Sciences, Business and Management, Decision Science and Multidisciplinary Sciences, as the obtained papers from these fields covered our research focus well, while those from other fields (e.g.Engineering) did not.Further, we decided to include only journal articles and conference papers published in peer-reviewed outlets and articles published in English, which is generally considered the international academic language (Genç and Bada, 2010;Merli et al., 2018).We employed a chronological restriction to articles published between 2014 and 2022, as the evolution of drone research in business and management scholarly journals started around 2015 (Merkert and Bushell, 2020).The data was collected in the period 21-23 February 2022, and repeated on 27 October 2023 while we worked on the paper revision with the objective to update the literature reviewat that time with a chronological restriction to articles published between February 2022 and October 2023.We ran identical searches on Scopus and Web of Science using our search strings.
The search in 2022 yielded a total of 329 articles with 17 duplicates.We imported the abstracts of the remaining 312 articles and checked whether they satisfied the above defined inclusion criteria.We excluded articles that referred to 'drones', 'UAV', or 'UAS' only in passing but did not study their applications, discussed mainly technical implementation details, looked exclusively at drone applications in the past, or looked at applications today but with no focus on potential implications or other future perspectives.Reading the abstracts resulted in an initial selection of 49 articles.After reading the full remaining articles, only 14 of them were found relevant to our study.Another seven articles were identified from backward and forward citation snowballing (Kivimaa et al., 2019(Kivimaa et al., : 1064) ) and were all found relevant.Including them resulted in a final sample of 21 articles.The identical search in October 2023 according to the same procedure led to an addition of three articles, resulting in a total sample of 24 articles (Table 1).
We analysed these articles with a qualitative content analysis in which we involved two researchers to thematically code (Gibbs, 2018) what was mentioned regarding drone applications and focus areas.Further, we noted the methodological approaches used and who were investigated or directly involved in the studies, if any.Finally, we positioned the articles on an implementation timeline: Now; short-term up to ten years from now; and medium-term to longer-term more than ten years ahead (Nordlund, 2012;van Notten et al., 2002).
Almost half of the studies [1,2,3,5,7,8,11,12,16,18,21] describe more or multiple drone applications for different types of missions, e.g. transportation in healthcare (e.g.Johnson et al., 2021 [9]), wildlife monitoring and detection of poachers (Kwon et al., 2017 [11]).Thus, drones either are or are envisioned to be used in different contexts, as highlighted by Boucher (2016 [3]) and, especially, Tan et al. (2021 [21]) who investigated the level of public acceptance across residential, recreational, commercial and industrial contexts of use.They found that the latter "had the highest acceptance level, followed by recreational areas and commercial areas while residential areas had the lowest acceptance level" (ibid.: 1).Furthermore, what might affect the public acceptance levels are fears and concerns and perceived potential benefits (ibid.).Around one third of the studies [4,6,10,17,20,23,24] focus on drone transportation, including last-mile delivery (e.g.Yoo and Chankov, 2019 [23]).And one study by Sigala and Langhals (2020 [18]) focuses on military applications but includes insights that will be relevant also to the development of civil drone applications.

Research gaps
Overall, the reviewed scholarly articles focus on the present or exhibit mainly short-term time horizons <10 years from now (Nordlund, 2012), while medium-term to longer-term studies of drones in a future >10 years ahead are rare.This is problematic, as drones are only in the emergence phase, as defined by Rotolo et al. (2015), and we are therefore at present only seeing the contours of how they might impact people and society.Thus, there is a need for visiting the far future to enable a back-cast of potential consequences, good and bad, of deploying drones full-scale in the short to medium term.
Furthermore, the three studies focusing on the longer-term future of drones in 2040 or 2050, each focus on a specific field of application (military missions, blood delivery and last-mile delivery of a diversity of goods), whereas the mainly short-term oriented studies show that there is no one prescribed path for drone applications but rather a multiplicity of potential development trajectories.This signals that drones in the long term might be applicable in completely different contexts than have yet been imagined, i.e. the number and nature of potential drone use cases might expand if the distant future is studied more than what it has already been.
Many of the studies focus on certain areas related to how people think about drones, such as their concerns or attitudes towards the unmanned aerial vehicles.We further observe that a variety of both quantitative and qualitative methods have been used across the studies.There is, however, a clear picture of mainly experts on either drones or the field(s) of application having been consulted.This is also the case for two of the three studies focusing on the longer-term future of drones, and the third of these studies presents scenarios developed by the authors themselves.Thus, none of the studies focusing on the longer-term involve citizen perspectives.To the best of our knowledge, we therefore conclude that no citizen visions of drones in the long term have yet been reported in academic literature.
Finally, we observe that the studies that did involve citizen perspectives focused on the short-term future and were conducted on different continents, i.e. in Asia [4,6,21], Europe [3,23] and the USA [2].

Methods
We chose Denmark as the context for our empirical study.For more than twenty years, Denmark has pursued an international innovation policy with clusters, and this with great success (Marti et al., 2022).It is not surprising, then, that in 2016 the Danish government presented the first strategy for the civil use of drones, allocating DKK 80 million to support industrial and research partners and their collaboration with the international community (Regeringen, 2016).This spurred the rise of several drone-oriented projects focusing on developing drone-related technology and breaking down the barriers for using drones in society.However, like in most other European countries, drone regulations still prevent the free use of drones, especially flights beyond the operator's visual line of sight (Frederiksen Harbo et al., 2019;Frederiksen et al., 2020).Thus, although Denmark has built up an expertise in drones, they are not yet commonplace in the air, meaning that many Danish citizens have not been directly exposed to them.This provided us with the opportunity to engage citizens with different levels of knowledge about drones, from novices to experts.Due to the focus on gathering citizens' far-future envisioning about the uses and impacts of an emerging technology, our research design required a social constructionist approach, i.e. a prospective investigation "principally concerned with explicating the process by which people come to describe, explain, or otherwise account for the world (including themselves) in which they live" (Gergen, 1985: 266) or, in our case, will live.This and the general lack of insights into citizen's farfuture perspectives on drones required an explorative research approach (Flick, 2014).It incorporated five phases: sampling, data collection, data analysis, data validation and data comparison.This research design enabled us to identify and conclude on what insights citizens' far-future visions of drone uses and impacts can offer to policy decision-makers.
In more detail, we collected data by means of several workshops where citizens wrote fictional stories about drones in the distant future of 2057.We chose this approach because earlier research suggests that to relationally negotiate and construct the world, humans use narratives (Sparkes and Smith, 2008).Narratives about the future are fiction stories.Fiction stories were also found serving as a means to gain access to self-transcending knowledge, i.e. visionary, not-yet-embodied tacit knowledge (Scharmer, 2001;Porter et al., 2004;Wolf et al., 2018Wolf et al., , 2022Wolf et al., , 2024;;Bajde et al., 2022) that includes "people's desires and ways to meet those desires" (Burnam-Fink, 2015: 52; see also Clarke, 1991;Schwarz, 2015;Bina et al., 2017).Specifically, we asked the workshop participants to write so-called flash fiction stories (Shapard, 2012), in the sense of very short science fiction stories.And science fiction according to Steinmüller (2010: 51) "comes close to a kind of fictional technology assessment".Science fiction stories are creative methods (Porter et al., 2004) that can deliver inspiration and new ideas for possible technology futures (Torras, 2020: 3) in the longer term (Miller and Bennett, 2008), facilitate experimental thought processes (Hauptman and Steinmüller, 2018) and diminish thought barriers (Steinmüller, 2010).Stories transport the writers' visionary ideas and thoughts, both good and bad (Schäfer, 2014) and help to explore fears, desires, catastrophes, changes and daily life adhering to technology in the future (Steinmüller, 1992).Johnson (2011) goes as far as terming the outcome of writing science fiction stories as 'science fiction prototyping', although, according to Bell et al. (2013), a typical, tangible prototype is not created.Johnson (2011: v) advocates that these science fiction prototypes can be "used as a step or input in the development process", while Graham (2013) thinks that these findings could be used as a possible input for regulatory or policy decision-making processes.
The described approach for the data collection thus served well our aim of gaining nuanced insights into how citizens nowadays envision the uses and impacts of drones in a distant future lifeworld.The next subsections describe the details of our methodological approach in the different phases.

Sampling
It has been shown that experts are very often involved in studies about the future (Schatzmann et al., 2013), which is also the case in the articles we reviewed in the previous chapter.However, experts may have cognitive limitations (Tetlock, 2017;Pang, 2010) but difficulties in admitting it (Fischhoff et al., 1982).Furthermore, Goldstein and Gigerenzer (2009) argue that experts are less powerful in foresight processes than non-experts, and Pang (2010) adds that experts' in-depth knowledge of a particular subject area does not make them good forecasters, or sometimes their forecast may be seen as random.Bonaccorsi et al. (2020) therefore suggest combining the opinions of experts and non-experts to overcome these biases.Adding to this, Gudowsky et al. (2012) say that the public is best placed to decide what is in its interest, and Fischhoff et al. (1982) believe that citizens should not be seen as misinformed.Based on this, we chose citizens with different levels of knowledge about drones for envisioning the future with the unmanned aerial vehicles.
For the flash fiction writing workshops, the sample composition reflected our aim to gain access to the broadest possible variety of citizen visions on drones in the far future.We thus decided to use a purposive high variety sample (Flick, 2014).We selected citizens in a way that the final sample would cover a broad age range as well as difference in genders and drone technology knowledge levels, as scholarly research suggests that diversification in these criteria will allow us to account "for the multiple perspectives that already exist in the present" (Zaidi, 2019: 22).This resulted in a total of 135 participants, who each attended one out of seven creative writing workshops that took place in the southern region of Denmark between November 2018 and March 2020.Table provides an overview of the workshops and participants.Furthermore, characteristics of the final writer sample can be seen in Figs. 1, 2 and 3.The level of expertise is based on the writers' self-assessment on a scale from 1 to 6 before the workshops started.

Data collection
We asked each workshop participant to write three very short flash fiction stories of around 300 words.The writers were instructed to focus on one topic or conflict and one person (Gaffney, 2012;Wolf et al., 2018).They were given only two to three minutes per story, so that they had no time to rationalise and access explicit instead of self-transcending knowledge while writing (Rosch, 2008;Ciolfi and Lockley, 2019;Wolf et al., 2022).Furthermore, they were asked to think of the year 2057, as writing about a remote future helps writers to detach from recent circumstances (Schutz, 1967;Mische, 2009).In total, we collected flash fiction stories about drones in the distant future.
The writers were first briefly introduced to contemporary designs and recent applications of drones.Then, they were informed about the research project, the planned use of the stories in case they decided to hand them in and options for remaining anonymous.They were also briefed that the stories would be written by hand, because the kinaesthetic experience of writing longhand affects both the quality and content of the outcome, it is conducive to creativity and longhand is difficult to edit while flow writing (Manwaring, 2020).
To get the writers started, three warm-up sessions served to get them used to the format of writing flash fiction stories.First, they were presented with a creative writing prompt in the shape of an unfinished sentence and given 1 min to continue a story starting with this sentence.For the following two warm-up sessions, pictures were shown to the writers, and they were asked to write a story based on each picture in min.After the introduction and warm-up sessions, there were three creative writing sessions lasting 2½-3 min each, where the writers were asked to imagine and describe in Danish or English language the world with drones in the year 2057: • First session: a main character and a situation in which s/he will be exposed to drones in 2057.• Second session: a positive encounter between this main character and a drone in 2057.• Third session: a negative encounter in 2057.
Once all three stories were written, writers who volunteered read their stories out loud to others.This spurred small debates about the pros and cons of implementing drones for the purposes suggested in the stories.Reading stories aloud to each other and discussing certain aspects could be a possibility for 'play', according to Buxton (2007), in the sense of "verbal playing around" to "invite, suggest, and question" and help to find the relevant insights (Buxton, 2007: Chapter 27).However, this part of the workshops was not recorded and therefore not included in the analysis.The idea was merely to give the participants the opportunity to express further thoughts and round off the workshop.Most of the writers provided their contact information, demographic data and the self-assessment of their knowledge level about drone technology.
After the workshops, all stories were first scanned and then transcribed, numbered consecutively and those written in Danish were translated into English.Here are two original examples from workshop 1: I stood there, still, unable to help the man lying before me in a wheelchair comprised of his innards.The pressure was overwhelming, a sudden need to shed my emotional being and rid myself of these feelings, which imprison me, then it showed up like a robotic angel descending from the heavens, the surgery drone operated on the waning man and restored his strength.
The drones are like companions; everyone gets them at a certain age.Kids as young as 4 years old can be seen with them.Mia's view of them changed through the years.They were like nannies the first years, but with age they became more like assistants.They still may be called by the names kids came up with.

Data analysis
All stories were first coded by the three authors individually, using an open coding approach (Miles et al., 2014) and a process of inductive concept development (Gioia et al., 2012).The same three researchers engaged in a second round of cross-verifying the coding and ensuring intercoder agreement (Miles et al., 2014), taking as a starting point the first order concepts from the first round.For every first order concept, each coder independently identified stories that belonged to this concept.New concepts were allowed to be added, and suggestions for renaming or subsuming first order concepts could be made.
Next, stories that only one coder had suggested to represent a certain concept were re-checked by the other two coders and discussed in the team.The result of this verification step was that some of the codes were removed or combined with other first order concepts.We kept track of the changes (see Appendix D for examples of the coding).
From the coding process, 115 first order concepts emerged: 67 specific drone applications belonging to either a private, business or public sector context and 48 specific impacts of drone use.The first order drone application concepts were clustered into 27 drone use cases and the   specific impact concepts into 19 impact categories that further served as second-order themes.Then, 10 drone use categories and three impact types were identified as the aggregate dimensions in the data (for the overall data structure, see Figs. 4 and 5 in Section 4).Finally, the first order concepts were further analysed with a focus on the writers' age, gender and drone technology knowledge level to identify any tendencies of specific groups suggesting certain drone uses.

Data validation
To validate the coding of the flash fiction stories, member checks (Lincoln and Guba, 1985) were performed to obtain the feedback of respondents to our interpretation of the stories.We contacted nine of the writers.They were selected to be as representative as possible in terms of age, gender and level of drone technology expertise (Table 3).
We scheduled an individual validation session of about 1 h with eight of the selected writers.In these sessions, we presented the stories they had written and the coding tree with drone applications (Fig. 4 in Section 4).We asked them if they thought it looked logical, and if we had coded their stories the right way.The writers validated our coding, and a couple of them provided nuances to their stories regarding the context of use of their envisioned drone applications.These few additional contexts were added to the first order concepts in the coding tree and helped to refine and finalise interpretations.It was not possible to use the same kind of validation with one of the novices, who instead got the questions by email and answered back the same way.

Data comparison
We then compared the drone use cases identified in the academic literature and the flash fiction stories and recognised that a range of the drone applications mentioned in the flash fiction stories were not mentioned in the academic literature (Appendix A).Yet, it is known that analysing public discourse eventually helps to identify future visions that are different from those discussed in science (e.g.Kwon et al., 2017).We therefore decided to do a "cumulative validation" (Holzer et al., 2012:128) of the academic literature by February 2022 by reverseanalysing the appearance of our 'new' identified drone applications in the public discourse with a Google advanced search (Soliman et al., 2021).
There were additional reasons for conducting a Google literature search: (1) High-quality literature may not be published in the academic journals, because the authors are either not expected or under pressure to publish in these journals (Grayson and Gomersall, 2003), (2) the grey literature may be more contemporary than the academic literature (Adams et al., 2016;Pappas and Williams, 2011) and (3) it may also be that the grey literature is the result of validation of academic findings (Adams et al., 2017).The Google search for text (PDF) included the following document types of grey literature: discussion papers (Lawrence et al., 2014), publications from governmental agencies (Levin, 2014), policy documents (Conn et al., 2003), committee reports (Benzies et al., 2006), working papers (Lawrence et al., 2014), institutional reports, project reports, technical reports (Schöpfel and Farace, 2021) and PowerPoint presentations (Benzies et al., 2006).The search was carried out using a Google advanced search.For search strings, we used those mentioned in column D in Appendix B, which are based on the identified first-order concepts (Fig. 5; Section 4).We limited the search for further publications by using the following refinements: search word: report, last update: past year, file type: PDF, as we would otherwise receive an unmanageable amount of unfocussed data.We ignored offline, spam and advertisement web pages (ibid.: 165) and included only publications in English.We ran an additional Google advanced search for videos, and then did not use the refinements 'file type' and the search word 'report', but the same search strings as for reports (Appendix C, column D).As the wildcard search on Google advanced search is not possible with an asterisk, we used the full terms.We tried to accommodate for the effects of Google advanced search results depending on the (re)searchers search history as well as the location by having two researchers placed in two different European countries while they did the same search (Piasecki et al., 2018).The search results for all three search strategiesacademic literature, Google search for text (PDF), and Google search for videosare shown in Appendices A, B, and C.
Overall, we identified 41 documents and 143 videos that fulfilled the defined criteria of a literature search.To present the extensive Google advanced search results in a clear, comprehensible and reproductive way (Gusenbauer and Haddaway, 2020), the following categories have been used: (1) none (if no search results at all or the information 'no results found' or only sponsored links were displayed), (2) none relevant (if search results were displayed but their content was not relevant), ( 3) number (if relevant search results were only displayed on the first page, then the exact number was noted), ( 4) several (if relevant search results were not only displayed on the first page but also on the second page, then the term several was noted in addition to the exact number of search results on the first page).To understand whether the search results covered the topics we were looking for, the search terms (combination of different words) were also searched for within the documents.If the search terms were found in the document, but in fact at different places in the document and thus possibly not covering the topic, then the search terms were put in inverted commas to find exactly that search term and thus clarify whether the document was relevant to our search or not.To check whether the videos really relate to the topic we were looking for, the videos were watched in overview.

Findings and analysis
In the following, we list and briefly describe the content of the citizen-envisioned drone use categories, use cases and applicationsand, next, both positive, negative and no impacts that we identified based on the analysis of the flash fiction stories.

Imagined drone use cases
Across ten use categories, we identified 27 distinct far-future drone use cases, which in total rely on the future availability of 67 different drone applications or services (Fig. 4).
Transportation provision: This use category comprises two use cases that both include applications or services in private, business and public sector contexts: • Providing pick-up and carry services relies on drones capable of fetching and carrying items (e.g. from stores or from a home to a workplace); drones that deliver goods such as pizzas; drones that transport cargo such as spare parts for wind turbines; drones that supply homeless people with foods and goods; and drones that deliver blood and medicine to institutions that are part of the public health system.• Providing transportation of living beings includes privately owned drones that transport people and animals; drones owned by service providers that offer passenger transportation services for both humans and animals; and hospital or state-owned drones that act as ambulances or perform welfare-related services.
Housing provision: This use category includes one envisioned use case: • (Mobile) sheltering refers to a drone that serves as a home comparable to a houseboat, but in the air.It enables private people to remain flexible concerning the location of their home.
Assistance (on-demand for humans): Many participants imagined drones in conjunction with on-demand assistance for humans.We identified five different use cases in this category:    • Providing personal shopping assistance is a use case in the private context, where drones accompany and assist humans while they shop.• In the Searching and guiding/rescuing use case, we find drone applications in the private and public contexts: Drones that support private people with pathfinding or locating family members, and drones that assist public sector rescue teams in finding and rescuing victims in cases of disasters or accidents.• The use case Interacting with and orienting customers happens in the business context and includes services such as drones that act as waiters, company butlers or sales representatives.• Assisting in domestic tasks is a use case in the private context, where drones for example serve as an alarm clock, brush humans' teeth or provide them a showeror more generally act as house assistant, butler or (private) nurse.
• Substituting tasks at work is a use case where drones replace their private owners at work in case of illness.
Work execution: This is the use category with the most use cases.Other than in the assistance use category, the drones do not act ondemand.They are imagined as performing routine work tasks independently ofbut instructed and monitored byhumans.In this category, we identified seven use cases: • Providing safety and keeping law and order is associated with the private context and a 'Guard drone' application, as well as with the public context and 'Police drone' or 'Ticketing drone' applications.• The use case of Providing public space maintenance is also imagined in a public sector context, for which a 'Cleaner and maintenance drone' is envisioned.Human relationship replacement: In this use category, drones replace humans in a private context.In the two related use cases, drones are: • Supporting humans, where the service includes a 'Motivator/ Comforter/Entertainer drone'.• Providing relationships through either a 'Companion/Friend drone' or a 'Girl-, boy-or kid-friend drone', some of which are sometimes even involved in intimate interactions.

Illegal activity facilitation:
In this use category, two use cases were envisioned: • Enabling harassment in a private context, which relies on the availability of a 'Spying drone' and/or a 'Bullying drone'.• Supporting criminals in a business context, which requires drones that can perform activities related to 'Human trafficking/kidnapping', 'Drug dealing and delivery [of drugs]' and 'Blackmailing'.
Remote warfare facilitation: This use category comprises only a single use case: • Enabling military activities.The specific applications include a 'Hunter drone' and a 'Killer drone'.Whereas the usage of drones capable of hunting people is limited to a public sector context where they would appear in swarms, drones used for killing humans are envisioned in the private, business, and public sector contexts.For the latter, drones have the capacity to carry a bomb or chemical weapons.

Spare time activity facilitation: This use category comprises two private context use cases:
• Facilitating exercise with the help of a 'Sports drone' that for example acts as a partner in games, such as badminton.• Facilitating gaming/games, envisioned as happening with the help of a 'Competition drone' that competes against other drones or a 'Play drone' that acts as an airborne playground for kids.

Communication technology use facilitation:
This use category combines drones and communication technology.The use cases are threefold: • Providing pictures and footage, where a 'Camera drone' is used in a private context and a 'Photographer drone' in a business context, for example by taking and uploading pictures and footage for realtors.• Facilitating meetings, where a 'Hologram-making drone' supports private people in need of help during meetings.• Supporting media usage, where a 'Communication drone' assists business actors with setting up and conducting meetings and a 'Technician/information drone' supports actors in a public sector context in media usage, for example by setting up online teaching environments.
Human capability enhancement: The last use category comprises a private context use case: • Augmenting human perception and physical ability, where a 'Handraiser drone' can assist lazy students and a 'Sensory enhancer drone' can become part of the human body and allow humans to for example to see in the dark.

Most frequently mentioned use cases:
The most often imagined use cases are 'Delivery service drone' (mentioned in 67 stories) and 'Personal and animal transportation service drone' (mentioned in 39 stories), followed by 'House assistant drone' (20 stories), 'Ambulance drone' (16 stories), 'Medicine drone' (13 stories) and 'Fetch and carry drone' (11 stories).Overall, private-use service drones that provide pick-up and carry services, transportation of living beings and assistance in domestic tasks are dominating the stories.

Imagined impacts of implementing drones into society
Three types of impacts of drone implementation were identified: Positive, negative and no impact.11 impact categories with altogether 33 specific impacts belong to the negative impact type; seven impact categories with altogether 16 specific impacts belong to the positive impact type; and only one category and one specific impact belongs to the no-impact type (Fig. 5).
Negative impact type dimension: The negative impact type dimension comprises the following impact categories: • 'Creation of negative reactions' in private contexts, such as arguments between humans or aggressive actions against drones, for example when humans try to destroy them.• 'Negative consequences of malfunctions' that often span private, business and public sector contexts, as when: • technical problems such as loss of power or computer failures cause non-reliability of drone services, • animals, humans or material goods get injured or damaged in drone crashes and accidents, • drones lose (part) of their load, • or when humans need to correct or deal with malfunctions of drones such as wrongdoing, incorrect understanding of commands or flying to a wrong destination.
The impact of drones getting lost is only mentioned in private and business contexts, and the listed reasons are sometimes unknown, while at other times it is described how drones are eaten by animals or hijacked: • 'Crime and remote warfare become easier' -across all contexts, the writers imagine that this will lead to drones being used to chase and kill.In the private and the business contexts, it is furthermore envisioned to ease the use of drones for stealing, spying and blackmailing and programming them to lose stuff on purpose.• Private-context related 'negative consequences of delivery' of goods by drones, such as humans becoming lazy and obese and an increase in consumption due to the ease of to-the-sofa supply.imagine that drones will be used as an excuse for not taking on social roles such as being parents or taking care of elderly parents or friends.
In the stories, a described negative consequence of using drones is that humans communicate and see each other less, become isolated and feel lonely or unhappy.• Drones are a 'cause of dependency' of humans on technology.In the private context, the use of drones is assumed to impair human abilities, including social and emotional competences, while in the public sector context, the writers fear that drones decide, govern or take over the world.• In a private context, drones 'perform concealed data collection and analysis', such as using AI to anticipate people's behaviour and needs, watching people while it is unclear who controls the drone and less privacy.
Positive impact dimension: The dimension of positive impact type comprises the following impact categories: • "Positive consequences for the environment" -for the private and the public sector contexts, less pollution is an expected effect.In the private context, a mentioned side effect with a positive impact is that leftovers or dropped goods serve as food for animals.• A 'safer lifeworld' that is imagined as the result of using drones due to impacts such as less traffic and prevention of accidents in the private context, and a higher degree of security through monitoring drones in the public sector and private contexts.• Several private context positive impacts resulting from drones that 'create enjoyment'.These range from that humans and animals enjoy the benefits of drone usage, such as the ride or the view, to the positive aesthetic experience from seeing the elegance of flying drones.Drones are expected to create new, often-valued jobs or activities andthrough thatfeelings of pride and joy in jobs as well as experiences of job valuation due to drones taking over routine, low value or dull tasks.• 'Friendships emerge' -encompasses only one private context impact, namely that drones become friends of humans.• 'Inspire artists' -reports of one business context use case in which a drone is accidentally inspirational when it improves a sculpture through crashing into it.• Private context impacts of drone applications that 'ease life': Drones are expected to make it easier living in the outskirts and to visit each other by improving access to remote areas and increasing travel speed and comfort.They are generally assumed to create convenience in mobility for people with disabilities or those travelling the world in their drone home, and they provide the possibility to shop and get delivery 24/7.• Drones 'increase efficiency', such asacross all contextsfast transportation and delivery of goods, humans and animals and enabling cost savings through inexpensive services, lower required manpower and public expenditure.In the private context, an envisioned positive effect is to save and/or free up time earlier spent on routine or low value activities for other activities.
No impact dimension: The no impact dimension comprises only one impact category -'no disturbance' -with only one specific impact.In the story that belongs to this dimension, the impact of drones is characterised as non-existing, because they are invisible like internet cables.
Most frequently mentioned impacts: The most often mentioned negative impacts are crashes or malfunctions that were caused by technical non-reliability (45 stories), drone usage for criminal acts (21 stories), opportunities for surveillance while staying anonymous (20 stories), and sky pollution (18 stories).The most often imagined positive impacts are increased efficiency due to fast transportation (27 stories), the time that humans save and can use for other purposes (22 stories), and the joy that is created when flying with drones (21 stories).

Controlling for citizen characteristics
When qualitatively analysing the potential impact of the citizen characteristics that we controlled for during samplinggender, age and drone technology knowledge level (novice, knowledgeable, expert) -on the envisioned uses and impacts, we identified some clear tendencies of uses being imagined by individuals with certain demographic characteristics, while other visions were shared across demographic profiles: • For the often-mentioned use cases of Providing pick-up and carry service and Providing transportation of living beings, the level of knowledge about drone technology does not play a role.However, the age group of people below 30 and of males dominates.• Drones assisting in domestic tasks and providing safety and law and order are predominantly imagined by males below 30 years, but hardly ever by experts.• Healthcare-related use cases are mainly imagined by males, predominantly between 30 and 39 years old.• All spare time activity facilitation cases are imagined by male novices.
• Only novices envision the use cases from the Communication technology use facilitation use category.
• Human relationship replacement is imagined by both genders, but only male participants below 20 years imagine intimate love relationships with drones.• Only male writers came up with the supporting criminals use cases.All but one are experts.However, in the harassment use case, almost all the stories are written by novices of both genders.• We can see that experts imagine, in the same use category, more complex use cases, such as: • Building inspection (novices) vs. Construction or Roadwork drones (experts) • Girl/Boyfriend (novices) vs. self-made Kids-drone (experts) • Hand raiser (novices) vs. Sensory enhancer drone (experts) There are also visible tendencies with regards to specific impacts of drones: • Predominantly males look forward to the positive impacts on mobility and delivery in the impact category "Eases life".• Females below 30 years are more scared than men of inefficiencies arising from drone useyet at the same time, predominantly this group mentions an increase of efficiency as potential positive impacts.• The writers mention crashes and malfunctions more often than humans or animals being hurt.

What is new?
We investigated what new insights our study has provided by 1) comparing the drone use cases mentioned in the flash fiction stories with existing sources; 2) looking for examples of how citizen characteristics, drone use cases and contexts relate to the impacts mentioned in the flash fiction stories; and 3) categorising the findings into overall types of drone use cases according to what can be performed already today and what will require new technological developments.

Comparison of flash fiction findings with existing sources
We juxtaposed the drone use cases mentioned in the flash fiction stories with those mentioned in existing academic literature (from Section 2) or in grey literature or shown in videos.
Many of the drone applications mentioned in the existing academic literature reappear in our flash fiction stories, such as last-mile-delivery of food and commercial goods; transport of heavy materials and people; emergency response with medicine, defibrillator or organs; search and rescue of people; tracking criminals and animal populations; surveilling public places; monitoring traffic; reforestation; spraying with pesticides; monitoring and inspecting construction sites; enhancing human senses; supporting sport activities; and racing and entertaining.Thus, in many ways, our study confirms what is already known from academic literature about existing and potential fields of application for drones.However, there are some differences.On the one hand, some drone applications that are mentioned in the existing academic literature do not appear in the flash fiction stories, such as co-creation of graffiti art (Herdel et al., 2022); scanning for metals underground and monitoring of aquatic life (Lakshman and Ebenezer, 2021); firefighting (Tan et al., 2021;Lawson and Rajan, 2023); taking up accident scenes (Lawson and Rajan, 2023); inventory control in warehouses (Maghazei and Steinmann, 2020); large displays for entertainment or information in public spaces, or cinematographic tools (Herdel et al., 2022).On the other hand, our flash fiction stories suggest completely new potential drone applications (listed in Table 4).
The new drone applications in the flash fiction stories, compared to those found in the grey literature, include a homeless people supply drone; transport of heavy materials, people, and animals; interacting with and orienting customers; emergency response with medicine, defibrillator or organs; diagnostic and surgery drone; home drone; waiter drone; house assistant (indoor) drone; bee drone; transformer drone; work replacement drone in general; counselling/friend/motivator drone; supporting media usage and sport activities; supporting bullying, blackmailing, human trafficking and kidnapping; and enhancing human senses.
New drone applications that are mentioned in the flash fiction stories but do not appear in the identified videos include a homeless people supply drone; transport of heavy materials, people and animals; emergency response with medicine, defibrillator or organs; diagnostic and surgery drone; home drone; waiter drone; house assistant (indoor) drone; bee drone; transformer drone; counselling/friend/motivator drone; supporting media usage and sport activities; supporting bullying, blackmailing, human trafficking and kidnapping; and enhancing human senses.
Taken together, there are drone applications mentioned in the flash fiction stories that we found neither in existing academic and grey literature nor in videos.Table 4 lists these new drone applications and shows excerpts from some of the flash fiction stories.

Emergence of a complex picture
When we relate the citizen characteristics, the imagined drone use cases, their contexts and the imagined impacts from the flash fiction stories to each other, a complex picture emerges.Table 5 lists some examples.
What we can see is that in the different contexts, different envisioned drone use cases in the flash fiction stories were presented along with positive and negative impacts, so that a very rich and comprehensive picture of use scenarios emerges.Interestingly, many of the imagined negative impacts are opposed by positive impacts that fall into the same impact category, such as the fear of job loss on the one hand and the creation of new jobs and increased job value by using drones on the other hand.

Categorisation of citizen-envisioned far-future drone use cases
Seen from today's perspective, the different use cases will require different levels of technological maturity.Based on this, the citizenenvisioned drone applications can be categorised into 10 overall types of use cases that we have labelled according to what they perform.
Four of the overall use case types relate back to activities that drones can perform already today, although some technological improvements will generally be necessary for them to do so with sufficient efficiency: 1. 'Eye in the sky': Citizens are aware that, equipped with a camera, a drone can be used for detecting and registering tasksit can monitor, inspect, survey, or record things.An example is the 'Environmental monitoring drone'.2. 'Errand runner': Since a drone can carry things, it can be and is used as a 'Delivery service drone' and thereby acts as an errand runner who transports and brings someone things, or it even transports living beings.3. 'Enerver': Drones are known as having the capability to, in a negative way, detect and follow someone, i.e. a drone can survey or pursuealso in a private context, such as the 'Bullying drone'.4. 'Eraser': A drone that carries things can also release its payload with the purpose of hurting or killing someone like the 'Hunting drone' and the 'Killing drone'.
Further, our study identifies four drone use case types that are complex and would require the development of additionalmostly diagnostic and emotional response relatedtechnological functionalities before drones can perform them at a sufficient level: 5. 'Emergency handler': Based on a combination of the capabilities of the 'Eye in the sky' and the 'Errand runner', citizens expect drones to be capable of taking over several emergency service tasks, such as diagnostics and surgery, home nursing or search and rescue missions in the private and public sector context.Yet, all these tasks also involve a diagnostic part that drones cannot perform so far.6. 'Entertainer': Today, drones can collect and display information and footage.Beyond that, projecting holograms for communication, setting up communication environments like meetings without human help, creating competitions or entertaining people will require further capabilities.7. 'Enabler': Citizens envision drones to assist humans by taking over domestic and business tasks of very different types.While some of these are technically possible today, such as setting an alarm or inspecting a building, performing most of the tasks will require evaluation, judgement and adjustment, such as herding, road working, cleaningor even pulling the duvet away long enough for the person to wake up.8. 'Enforcer': Drones of this type are imagined to independently enforce law and order as guards or police members.Again, this includes evaluation and judgement activities that drones today cannot provide.
Finally, we find two use case types that drones cannot perform today: 9. 'Expander': These drones provide an expansion of the living space, be it for private housing or public use such as bridges or playgrounds.Inhabitable or flexible drones of that size are not portrayed in the recent scientific literature.10. 'Engager': Drones that provide psychological help or counselling, companionship and act as a girlfriend or boyfriend are not impossible in general.Yet, drones are today limited in the way they can make sense of situations, become emotionally engaged with humans, and respond in a meaningful way to inquiries for advice and help.Acting as an 'Engager' will require that drones execute tasks or empathise with humans, which calls for advanced motor and cognitive skills of drones in the future.

Discussion
We have investigated different citizens' visions of drone uses and impacts in the distant future, as we believe that these insights have something to offer to policy decision-makers residing at both governmental and corporate level (Brick et al., 2018).
The citizens represent those who will potentially be overflown if drones become commonplace in societyand, to the best of our knowledge, our flash fiction study is the first to directly factor in the voice of different citizens on how they imagine the society with drones in a far future of 2057.Existing academic literature involving citizen perspectives has investigated mainly short-term time horizons <10 years from now (e.g.Ganjipour and Edrisi, 2023;Hwang et al., 2021a;Yaprak et al., 2021), whereas the few studies we identified that focused on the longer-term future in 2040 or 2050 did not involve citizens and each focused on a specific field of application, i.e. military missions (Sigala and Langhals, 2020), blood delivery (Sodero and Rackham, 2020) and last-mile delivery of diverse goods (Peppel et al., 2022).
Our study underlines that there will expectedly be a multiplicity of potential drone development trajectories in various fields of application.Some of these may be spurred by extensive technological development in terms of higher levels of flight autonomy.However, an expansion of both motor and cognitive skills of drones will be needed.Furthermore, our study shows a range of potentialboth positive and negativeimpacts that policy decision-makers should consider before allowing or preventing the further implementation of drones in society.Such considerations are relevant at both governmental and corporate level, as both types of decision-makers should take on the responsibility for a sound, sustainable implementation of emerging technology (Spiekermann, 2023).

A citizen-specific nuanced picture
Overall, our study confirms that citizens imagine drone use cases that are discussed in existing literature for the soon-to-come future to be still operating in the far future, such as delivery drones (e.g.Hwang et al., 2021aHwang et al., , 2021b;;Ganjipour and Edrisi, 2023).However, their visions go  beyond what could be expected in many ways, thereby providing a nuanced and citizen-specific perspective that provides insights for policy decision-makers in three different ways: First, we have identified some truly original ideas that represent completely new drone uses, such as the Transformer drone that can take any shape desired and become, e.g. a bridge in the public space and the Shower drone that can bathe people in their private home.Other use cases that we identified extend existing ones through introducing them into alternative contexts, e.g. when kids in the future use a drone for hunting other kids in a bullying way or when a drone becomes a privately-owned security guard.Further, citizens imagine drones as substituting human work as well as being used in their private sphereand even engaging closely with humans.Other parts of the stories point to more radical uses of drones than what is already known or envisioned in the academic literature.As an example, instead of simply delivering food to people as scientists would expect today (Hwang et al., 2021a(Hwang et al., , 2021b)), a drone can in the far future act as a Waiter drone at restaurants, where it does not only deliver the food to the table but also has a conversation with the visitors.Another example of a more radical use is the Defibrillator drone that cannot only deliver a defibrillator to where a person suffering a heart attack is locatedsomething that is possible today (Koshta et al., 2022) -but will also set up the defibrillator by itself and perform cardiac massage on the patient, i.e. without the need of human assistance.Hence, the collected flash fiction stories extend the perspective that is available to policy decision-makers with a citizenspecific perspective about imagined drone applications in the future.
Second, our study also allows us to contrast the frequency of citizens imagining specific use case domains with the frequency in academic literature.For example, Herdel et al. (2022: 7) identify the domains of 'Emergency', 'Entertainment', 'Communication' and 'Sports' as the four most mentioned domains of recent and future drone uses in the Engineering literature.Instead, our citizens envisioned most of all use cases in the domains of 'Help/Assistance' which only occupies the fifth place in the Engineering literature, and specifically assistance in the form of pick-up and carry service and transportation provision and in domestic tasks.Other than the engineering perspective that focused on outdoor uses of drones (ibid: 13), almost half of the flash fiction stories include envisioned drone uses inside (mostly private) houses.Such insights can support policy decision-makers in prioritising citizen-desired drone use cases over those imagined by engineers from a feasibility perspective, and thereby advance the conditions for future use cases that may contribute to the wellbeing of humans and society (Spiekermann, 2023).
Third, our study has revealed ten general types of use cases that we labelled according to what drones are imagined performing, ranging from an 'Eye in the sky' to an 'Engager'.Some of these use cases are already possible, while others will require technological developments.The labelling will be helpful for policy decision-makers, as earlier research indicates that speaking the language of the public is important for mediating dialogue across private, business and public sector contexts about future applications (e.g.Fung, 2015).Further, the feasibility classification allows policy decision-makers to assess the technological and societal readiness level together with the scale of political challenges that need to be dealt with before certain use case types can become reality.For example, the 'Expander' requires not just technological innovation to increase the inhabitable or usable space, but also legal changes that would regulate inhabiting the airspace.In that sense, we confirm that policy decision-makers can gain actionable insights that can be mapped in time from early opportunity identification processes that involve citizens in emerging technology use case development (Buxton, 2007;Teece, 2017;Xing et al., 2022).

Points of attention for policy decision-makers
Ideally, "policy decisions must be transparently informed by evidence" (European Commission, 2021: 1).However, the impacts of starting to or increasingly implementing emerging technology remain uncertain and ambiguous until sometime in the future (Köhler and Som, 2014;Rotolo et al., 2015).Thus, the foundation for making decisions about emerging technology implementation will need to come from anticipations of potential impacts.
Our study provides both an overview of and details about the potential distant future with drones that can serve to inspire today's policy decision-making regarding the further implementation of this emerging technology.The different use cases provide new and nuanced insights pinpointing potential both positive and negative impacts of increasingly implementing drones in society (see Appendix E for suggestions to what policy decision-makers could consider).These impacts are envisioned by diverse citizens, including both novices and experts on drones, which highlights the relevance of anticipatory governance and listening to the voice of different stakeholders (e.g.Guston, 2014;Lehoux et al., 2020).
What we also see from our study is that quite a few of the imagined impacts are in opposition to each other.For example, on the one hand, the increased presence of drones might lead to visible and audible 'sky pollution' but, on the other hand, drones might be the most climatefriendly means of mobility (Table 6).This represents a paradox that policy decision-makers must deal with, i.e. consider if the advantages outweigh the disadvantages and if anything can be done to minimise the latter, of course.
Paradoxes are evident also in other complex technology fields, such as artificial intelligence (Bremmer and Suleyman, 2023), where it is also necessary to embrace the opportunities to gain the advantages but simultaneously consider how to do it to avoidor at least tamethe disadvantages.Paradoxes are characterised as "contradictory yet interrelated elements that exist simultaneously and persist over time" (Smith and Lewis, 2011: 382).Thus, policy decision-makers cannot eliminate tensions but must attend to competing demands (ibid.: 381).This underlines the need for listening to different stakeholders but also for using different and perhaps a combination of instruments.For example, based on the underlying logic that it can be fatal to living beings if autonomous drones malfunction, the goal would be to avoid malfunctioning drones in the first place.Governmental policy decision-makers might therefore adopt primarily regulative but also economic instruments to force businesses to develop safe and trustworthy solutions andif things go wrongpunish them by means of imprisonment and/or fines.However, over-regulation could stifle the necessary technological progress (Vargo et al., 2020), as businessessuch as those developing, manufacturing, and/or commercialising drone solutions or those providing critical infrastructure such as 5G network or unmanned traffic management systemswould hesitate to fully deploy their solutions.Should they

Table 6
Examples of opposing negative and positive impacts that constitute paradoxes.

Examples of potential negative impacts
Examples of potential positive impacts • Malfunctioning drones may cause damage to material goods and injury of living beings.
• An increased use of aerial drones may entail less ground traffic and accidents.
• Lost drones and payloads may fall into the wrong hands and/or never be found.
• Lost payload may be consumed by and thus benefit those on the ground who find it.• Omnipresent drones may create both visible and audible sky pollution.
• Electric-driven drones may pollute less than other means of mobility.• People may be deskilled or lose their jobs because they are supplemented or replaced by drones.
• People may feel pride and joy from working (together) with drones and avoid "low value", dull tasks.• People may become obese, communicate less, and feel increasingly lonely if drones deliver everything they need.
• Peoplealso those with disabilitiesmay find it easier to get hold of things and reach each other and thus settle evenly throughout the country.• The combination of drones and artificial intelligence may lead to hidden data collection and analysis and thereby less privacy.
• People may enjoy a safer and easier life with a higher degree of security and comfort if drones by means of artificial intelligence generate and use data to warn, inform, devise, advise etc.
M.H. Frederiksen et al. choose to do so anyway, they must focus on public affairs and have a corporate governance system in place with instruments that can minimise potential reputational and economic damages or punishments in case their drone solutions malfunction.The paradox of wanting drone solutions but not really allowing/deploying them unless they are 100 % safe means that decision-makers at both governmental and corporate level face a demanding ride where they must alternately loosen and step on the brake to enable sustainable innovation on wings.Governance of paradoxes is thus a balancing act, which is further complicated by the fact that technologies rarely operate in isolation but rather in conjunction with each other (Bailey, 2022).Furthermore, as technology changes, the paradoxes may very well also change or occur over time.For example, to begin with, the creation of new traffic layers and increasing deployment of aerial drones may lead to less traffic congestion on the ground.However, over time, the air might be crowded with a multitude of both passenger and cargo drones, which could then entail a queue in the air (just like some of the flash fiction stories mention).Thus, policy decision-making concerning drones must be dynamic and adjust in pace with technological developments and the emergence of the societal consequences of past decisions.
Taken together, we contribute to literature on policy decisionmaking and anticipatory governance.The drone use cases identified via our study reveal both positive and negative impacts that the citizens envision, thus allowing for the emergence of a complex and nuanced picture that links drone use cases to citizens' fears and hopes (see Table 5 in 4.4.2).This offers the opportunity for policy decision-makers to gain insights into the imagined future lifeworld in which these drone cases are envisioned and put themselves in the shoes of different citizens.Furthermore, the often-opposing impacts highlight the need for policy decision-makers to be able to deal with paradoxes, which complicate matters.
Our flash fiction stories can serve the purpose of discussion openersthey represent an invitation to policy decision-makers to take the citizen perspective in foresight and challenge their own imaginations of potential far-future consequences of implementing drones, as demanded by Buxton (2007).Governmental-level policy decision-makers will benefit from such an approach by extending their perspective of how the larger-scale implementation of drones will potentially alter our society.They can thus better identify appropriate instruments for regulating or supporting initiatives through back-casting towards citizen-desired futures, which should be updated regularly, as "the desired future is not determined in advance of the analysis but is an emergent property of the process of engaging with users [...] a process of social learning about possible and desirable futures" (Robinson, 2003: 839).Likewise, corporate-level policy decision-makers can use insights from backcasting towards desirable futures to decide on the development of sound, sustainable and wanted drone solutions (Quist and Vergragt, 2006;Spiekermann, 2023).

Limitations and future research avenues
Our study identified a range of potential far-future drone uses and impacts of an increased implementation of drones in society.We are aware that the qualitative nature of the flash fiction study does not bring evidence to policy decision-makers and that we are trying to shed light on something that is per se uncertain and ambiguous.Furthermore, we realise the Collingridge-effect describing "a dilemma in which design inevitably comes before [...] assessment since the design influences heavily how the technology will interact with society" (Buckley et al., 2017: 54).What we suggest is a pathway for qualifying decisions and subsequent actions about emerging technologies such as drones that can with advantage be based on considerations of how those who will potentially be affected by technology deployment envision potential positive and negative consequences hereof.
We have identified paradoxes concerning drones that policy decision-makers at both governmental and corporate level must tackle.These paradoxes are also to be regarded as propositions for what to consider and investigate in the years to come when drones become more frequently used platforms.Thus, we propose that future research should focus on what would be the appropriate and needed decisions and actions to ensure that negative impacts can be avoided to the best possible extent and positive impacts can be enabled through the right set of incentives etc.Furthermore, studies considering citizen visions should be conducted continuously to monitor and identify emerging patterns as well as anticipations, perceptions and visions among those who will be overflown by drones as they become commonplace in society.We limited our study to data that we collected from citizens residing in Denmark.Although our sample consists of citizens of different nationalities, future studies are needed to expand this geographical limitation, i.e. as citizens of other countries may associate differently with drones.And future research can focus on using the same methodological approach as we did but studying a different technology so that it becomes possible to compare cases.
Finally, a possible next step could be to investigate how the identified far-future use cases already today can be used by firms in a meaningful way for the development of innovative products and services that can be sustainable in the long run.Ute Klotz is Professor of Business Information Technology at the Lucerne School of Computer Science and Information Technology, Lucerne University of Applied Sciences and Arts in Switzerland.She is also Co-Head of the focus group 'Technologies for the Digitized Working Environment of the Future' in the interdisciplinary thematic cluster 'Digital Transformation of the Working Environment' at LUASA.Her research focuses on future studies, the interfaces between humans, work and technology and methods that make the future more tangible and predictable.https://www.hslu.ch/en/lucerne-university-of-applied-sciences-and-arts/about-us/people-finder/profile/?pid=228.
M.H.Frederiksen et al.
M.H.Frederiksen et al.
CRediT authorship contribution statement Marianne Harbo Frederiksen: Writingreview & editing, Writing original draft, Visualization, Validation, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization.Patricia Wolf: Writingreview & editing, Writingoriginal draft, Validation, Methodology, Investigation, Formal analysis, Data curation, Conceptualization.Ute Klotz: Writingreview & editing, Writingoriginal draft, Validation, Methodology, Formal analysis, Data curation, Conceptualization.Patricia Wolf, PhD, is Professor of Integrative Innovation Management at the Centre for Integrative Innovation Management at the Department of Business & Management, University of Southern Denmark.She is furthermore Private Lecturer at the Department of Management, Technology and Economics at ETH Zurich, Professor of Innovation Management and Future Studies at Lucerne University of Applied Sciences and Arts and member of the universities' Future Laboratory CreaLab.Her research focuses on collective envisioning processes as catalysts for imagining the impact of trends and emergent technology adoption in society.https://portal.findresearcher.sdu.dk/en/persons/pawo.

Table 2
Overview of the flash fiction story writing workshops and participants.

Table 3
Overview of the writers selected for the member checks.Despite also being used in a public sector context, the uses imagined as part of the use case Providing emergency healthcare differ from the emergency call drone application.Providing emergency healthcare includes a 'Defibrillator drone', a 'Diagnostic and surgery drone' and a 'Counselling drone', which are drones that can provide actual medical or psychological services.

•
In a private context, 'cause of job loss' relates to that people lose their jobs for various reasons: Because they are replaced by drones, lack the necessary skills for a job as they are not taught anymore (e.g.driving a car) or are illiterate in operating drones.tobe resource and time-consuming.The writers also describe that even though drones give people more leisure time, people keep being busy.•Drones can 'cause environmental problems', such as sky pollution and queues of visibly and audibly disturbing drones in the private context and an envisioned shortage of materials that are used for building drones in the public sector context.• 'Increase of human isolation' in the private context: The writers • Drones 'cause feelings of anxiety' in the private context.These include the fear of crashes or malfunctions, include the recognition of own drone illiteracy or emerge because drones are perceived as scary black boxes that are impersonal, unpredictable and have no feelings.•'Drone usage is cost and time-consuming', leading to inefficiency in the private context.Additional costs associated with drone use are insurances, transport and delivery service charges.Maintenance and repair, such as setting drones up and following them on their mission, are expected

Table 4
New use cases that appeared from the flash fiction study.Excerpts are from the flash fiction stories.

Table 4
(continued )Some people still go to the gym, but even with drones doing so much of the work some people still seem too busy for that.As if that extra time isn't extra anymore.(22-year-old Danish female; novice)

Table 5
Examples of the complexity of the flash fiction stories.