Although making a success of smart farming is already tricky and adding new complications might appear unhelpful, the task must be to recognize and confront the complexities, rather than sidestep or ignore them. The question is how to construct smart farming innovation processes that yield more effective configurations?To avoid producing misconfigured innovations, smart farming requires a reimagined innovation process. I argue the challenge is to imagine and realize an innovation process that can yield ‘emancipatory smart farming.’ Some clues of what such an innovation process might entail are provided by research on the possibilities of pursuing ‘responsible research and innovation’ in smart farming. With a focus on anticipation, inclusion, reflexivity, and responsiveness , RRI tries to respond to the new “socio-ethical dilemmas” called forth by contemporary technological developments. For example, as stated by Rose and Chilvers , “[i]n the rush to embrace smart agri-tech, we are in danger of forgetting the wider network of other innovations that play an important role, but may also affect societies in different ways” . One focal point of RRI is therefore “to stage reasoned deliberations on technological needs and concerns between historically marginalized food system actors and prominent decision makers in government” . An ambition is that RRI might become “a rubric for guiding innovation toward socially and ethically acceptable ends” . At issue is examining “interrelations between multiple co-existing innovations in sustainable agriculture [to] promote the cultivation of distributed responsibilities across wider innovation ecologies” . In the New Zealand dairy sector, for instance, RRI has recognized that smart farming will yield “adapted advisory structures, potentially leading to displaced farm staff and service providers” . Moreover, shifts associated with smart farming technologies might have a “major impact on the cultural fabric of what it means to be a farmer,” in part because they can entail “detailed monitoring by agricultural equipment makers, input suppliers, u planting gutter processors and retailers” . There are reasons to applaud RRI. It signals at least an interest in trying to integrate societal concerns in technical developments; and opens avenues for new engagements between groups that might not otherwise interact.
It is a close approximation of what an appropriate topological repertoire might look like because it emphasizes the visibility of stakeholders, actors, and material realities that otherwise can be marginalized or ignored. However, RRI in smart farming still fails to produce adequate configurations. It operates via a misplaced insistence that agricultural innovation can successfully reconfigure sociotechnical relations in one domain, without also pursuing systemic or structural change. In short, it is necessary to continue insisting on the need for reimagined smart farming innovation processes that work to sidestep the misconfigured innovations evident in today’s smart farming developments. A pertinent example of what might be possible here is the development of farm OS, which draws on activist engagements and explores how smart technologies can empower communities, through actions of solidarity and co-learning . The software helps farmers record, plan, and manage their operations. It is open-source, produced under a general public license, and is easily hackable, in contrast to proprietary farm management software. Farmers can integrate diverse tools, such as drones for capturing aerial imagery or sensors to record temperatures, and thereby retain latitude to configure their operations in astute ways. In its effort to unsettle established smart farming structures and enable farmers to take back control over the software and data they produce, farmOS resembles other efforts to hack and repair farm technologies . It also reflects a much wider societal shift whereby activists, community groups, or others in civil society embrace contemporary technologies and take advantage of the emergent affordances to pursue “productive resistance” . A key dynamic of digital life today is growing realization that ‘smart’ use of software platforms requires re-platformizing society so urban citizens as much as rural farmers can take advantage of technological affordances without reproducing a platform economy dominated by a few enormous firms . It is therefore illuminating that farmOS is part of a new partnership called Open TEAM that aims to create a platform to facilitate “soil health management for farms of all scales, geographies and production systems” . There is scope today for farmers and connected others to overcome the problems of ‘actually existing’ smart farming and the misconfigured innovations it churns out. ise from new possibilities on the technological horizon. Hitherto, a technical limitation on smart farming developments pertains to the uneven roll-out of high-speed internet access between urban and rural areas . But there is now evidence that 5G networking technologies using TV White Space or ‘frugal 5G’ could be a ‘game changer’ for rural Internet access. If there is to be a ‘what next?’ of smart farming, it will build on what we find actually existing today to create new possibilities for embedding food production within the wider ‘planetary cognitive ecology’ , with unpredictable outcomes.
One relevant near-term scenario emerges from research on the ‘internet of people’ , a term used by computer science researchers with a view to building on and improving the relatively passive ‘internet of things.’ In the “Next Generation Internet” they are exploring, the internet of things is not swept away but rather a “new reference architecture” is carefully-crafted onto it with a view to overcoming problematic features of the “current-Internet data-management paradigm [such as] constant monitoring of users’ behavior by global platforms to provide to them ‘navigation’ and filtering services to find relevant data embedded in the huge amount of available data” . The overall design calls for a “human-centric perspective” at the scale of implementation and a novel “data-management Internet paradigm” in which devices are proxies of humans and constantly exist in context and operate in self-organizing networks that create new efficiencies because the need for human decisions is minimized. Significant features include use of new 5G capabilities that enable relatively autonomous ‘device-to-device’ communications across ‘pervasive communities’ of connected users. Per an IoP manifesto , devices are designed to ‘be social,’ ‘be personalized,’ ‘be proactive,’ and ‘be predictable.’ The underlying notion is that the IoP will use new arrangements and practices to engender economic efficiencies and positive social impacts. But there is every reason to expect the types of behaviours and interactions proposed by this line of research to impact on ‘smart farming’ practices. Consider a hypothetical example of how the IoP might operate, which, in the absence of available real-world examples to use, I adapt from contributions to the IoP literature : Maxine is a dairy farmer and cheese producer. Her cheese sales are disappointing. She’s confused and worried. She searches online for new recipes. Her phone knows a new recipe or idea is needed [‘be personalized’]. It shares this info with devices belonging to Maxine’s friends [’be social’]. At a social event soon thereafter, a phone belonging to a friend of Maxine overhears2 someone called Sandy say that Maxine’s cheese reminds them of another cheese they ate on holiday in Holland. It sends Maxine’s device a message along these lines [’be social’]; shares the ingredients and recipe of the Dutch cheese [’be predictable’]; and suggests a tweaked production process [’be proactive’]. Maxine’s device also communicates with quasi-autonomous devices in the cheese cellar [’be social’] to produce a new test batch. Some months later, Maxine has produced the new cheese product. Her device then detects that Sandy will be nearby soon [’be social’] and arranges for a sample pack to be delivered to her [’be proactive’]. At the same time, Maxine’s phone arranges for sample packs to be sent to other people who match Sandy’s sociological profile [’be proactive’]. Their devices respond to say they like Maxine’s new cheese and Maxine’s phone sends them discount coupons for their next purchase [’be predictable’]. Today’s smart farming developments lay the ground for emerging operations in the IoP: devices such as phones, or sensors to measure soil moisture or temperatures, are now operating on farms all over the world; software platforms are integrating actions, collecting and analysing data, and providing pertinent information to guide decisions; and autonomous machines are already in action.
All of these arrangements of devices and sensors share information according to protocols and standards worked out in the context of today’s technological limits and possibilities. The scene is therefore set for new rounds of investment in technologies that adapt architectures and yield realities like those posed above. As such, tomorrow’s protocols and possibilities will build on the normality of devices and sensors contributing to on-farm intelligence and efficiency but with a view to delivering results impossible hitherto. As suggested by Maxine’s case, then, smart farming in the IoP still relies on human intelligence but the abilities of her farm operation to survive is upgraded and amplified by protocols and standards that grant proxy devices autonomy and intelligence to proactively prompt new connectivities and relations. Maxine’s relations with others are mediated, filtered, and ranked; her digital life draws on new affordances developing dynamically within pervasive communities operating across a proactive internet. Beyond notions of the ‘nanny state’ infusing debates about communitarian governmental action, the IoP scenario is more akin to people living with numerous devices acting like ‘guardian angels,’ planting gutter with autonomous device-to-device decision making based on assumptions about the needs and possibly the desires of the individuals ‘they’ oversee. Maxine may be conscious of decisions she makes to engage the internet and might even understand or be sent information about autonomous device-to-device activities pinging messages and moving data according to underlying protocols; but much of her social life in the IoP also unfolds without her active participation. It is a new rural scene; an image of a different society from today’s, not least because it suggests the arrival of a new cognitive ecology underpinned and driven by AI, with social relations played out in numerous colliding “regions of technical autonomy” . Taking stock of the IoP scenario, there is clearly a strong possibility that smart farming in this forthcoming context will unfold via further rounds of misconfigured innovations. The dynamics of capitalist accumulation will no doubt pervade the design of protocols, devices, and services. Per the orientation of practices within so-called “surveillance capitalism” , tech firms such as Google or Amazon – as well as agri-TNCs, with their new data science profiles – will explore opportunities in the IoP to construct a more predictable world. Their challenge will be trying to contain the latent capacities and chaos of human and non-human action within tight profitable parameters; to thereby reduce the scope for uncertainty and contingency to interrupt flows of decisions informed by populations of sensors laid out and communicating with each other according to algorithmic models of society.
However, the objective reality of space is that no computational architecture can make sufficient calculations to overcome the inherent and pervasive “chance of space” . Maxine operates in a contingency-laden context before any IoP devices arrive on the scene. What happens with digital technologies in general, and the IoP in particular, is simply now that “the chance of space swells” . In the IoP, Maxine engages a new rural scene that amplifies chaos with unpredictable outcomes. As such, unexpected dynamics might come to the fore. The distinction here is between the architecture of digital life and the actual lived experience of digital subjects, which always entails “intersections and recursive relationships” playing out via “iterative interplays” . As evidence from research on digital worlds demonstrates – and as I have discussed via reference to developments such as farmOS – contemporary and emerging digital devices and services provide affordances for subjects to use technology in unexpected ways, including for the sake of resisting oppressive social formations. In rural space, smart farming seen through the lens of IoP research might place new value on intelligent, efficient, and even to some extent ruthless practices that squeeze as much profit from land and labour. Nevertheless, and emphatically, outcomes of the technological shifts at play here remain unwritten. Like farmOS or other efforts to re-imagine smart farming technologies, the IoP might create scope for users to create new forms of cooperation, reciprocity, and solidarity. There is significant scope for further investigation into this emerging scene.