When Things Shape Values (1/3): Data and Network Architectures in the Internet of Things

This blog post is the first of a three-part series looking at the ethical stakes of different technical elements that make up the Internet of Things (IoT). Here I discuss data and network architectures, and in the next posts, I will touch upon hardware components, energy, maintenance, repair, and interfaces.

Changing the focus: From Values to Things

A common way of studying the relationship between ethics and technology is to look at the shared values of the different groups involved in the process of design and development. This approach usually builds upon constructivist research focusing on how the assumptions and interests of actors shape the final product (Bijker et al., 2012). Echoing this approach, in Virt-EU we have argued that the European IoT landscape is marked by what Robin Mansell (2012) calls mainstream and alternative imaginaries of technology, with the former focusing on issues such as privacy and trust and the latter more interested in advancing transparency and openness.

More recently, a ‘material turn’ in academia has tended to emphasise the way the specific characteristics and affordances of devices, objects and things can also shape societal values (Marres, 2012). As Swierstpra and Rip put it, technology can “rob moral routines of their self-evident invisibility and turn them into topics for discussion, deliberation, modification, reassertion” (2007, p. 6). In the case of the Internet of Things, this means that building IoT devices or software can make more salient specific moral dilemmas or concerns, informing the morality of the groups involved in the process of design and development. As I explore here, the ethics that shape the IoT is the product of an interaction between human values and technical properties whose outcome is difficult to predict.

In the case of our research, this way of understanding ethics has two main advantages. The first one is that it allows us to move beyond the almost automatic association that designers and developers do between ‘ethics’ and ‘data’. Instead, when observing and talking about the process of design and development, it is easier to identify a more varied range of moral challenges that are not framed as ‘ethical’ by the actors involved. The second advantage of this approach is that it allows us to study ethics in the language of designers and programmers. By this, I mean that, instead of expecting them to verbally express their visions, we can see the process of coding, prototyping or business model development as a form ethical argumentation that does not rely on verbal argumentation but on technological ideation and practice.

Drawing on an inductive reading of the ethnographic observation and participatory methods carried out by researchers at the LSE, ITU, and CIID, in this blog series I identify some technical components that usually become ethically salient in the process of design and development of IoT devices. From now on I use the word ‘developers’ to refer to designers, developers and business managers of European IoT companies and organisations.

Data

Data is what usually first comes to the mind of developers when they are asked about ethics. Although ‘big data’ is usually framed in positive terms as a commercial opportunity, in practice the increasing velocity, volume, and variety of data seems to be a ‘big headache’ for people who have to decide what is going to be collected or how is it going to be archived, for example. As a result, and as a programmer told us, very often startups simply “don’t understand their data”. The implementation of the GDPR in Europe introduced an additional element of complexity. Under this scenario, startups have tended to respond in two ways: by ensuring the anonymisation of the information – an approach that does not necessarily solve all privacy issues – or by stopping all processes of data collection.

However, we have also found cases in which the dynamic between data and values have had a more productive outcome. Some developers do approach data with enthusiasm, taking advantage of its capacity to advance values such as transparency, openness and social justice. For example, Provenance and TagItSmart! are projects that seek to improve the process of consumption by allowing users to access additional information about the production and supply chain of products. Founded by Jessi Baker and based in London, Provenance employs blockchain to store the data and allows to interact with companies. TagItSmart!, on the other hand, is an EU’s Horizon 2020 project, and uses QR codes made with an ink sensitive to temperature and humidity.

These cases illustrate that the outcome of the interaction between technical affordances and ethical values is not always determined. One developer’s headache can be another’s resource for ethical tinkering, making possible the enactment of principles such as openness and transparency.

Network Architectures

The Internet of Things draws on the idea of ‘connected devices’ or ‘machine-to-machine communication’ In technical terms, realising this sociotechnical imaginary of connectivity requires the employment of cables, data centres and standards, among other resources. As usually happens with infrastructure, the technical, embedded and pervasive character of the network architecture of the IoT makes it harder to interrogate its ethical character. However, issues of ownership, compatibility, security and access are not morally neutral. The choice between 4G, Wi-Fi, LoRa, Bluetooth® mesh networking, Sigfoxor NB-IoT has an important technical component, but the decision will nonetheless have environmental, economic and societal consequences. Interestingly, in our fieldwork we have seen a proliferation of commercial and non-profit alternatives for connectivity that, by mobilising specific values, bring to the fore the non-purely-technical nature of the network supporting the Internet of Things.

Dowse, for example, is a home networking technology that, among other goals, seeks to make evident the ethical dimension of connectivity by purposely deleting military references such as ‘firewall’ and ‘defense’ of its documentation. Additionally, this project is being developed by the non-profit and open source organisation Dyne, raising issues of governance and ownership to the discussion.

In our research we have also found that alternative network architectures can inform community building. An analysis of the IoT Meetups in Europe reveals that, along with geography, the resources for connectivity employed to build devices is an important variable in the formation of sub-groups within the IoT community. An interesting focus for further research could be in the way the different technologies they use can shape their group values. Along these lines, in a Meetup in London, we heard from a developer that he uses Android Things not only because of its technical features but also because of the quality of its community, with people supporting and providing advice to each other in the process of development.

Although we usually tend to think of standards as related to computers, they have an important ethical component. As Star and Bowker argue, infrastructures “standardize both people and machines” (2002, p. 234–235), which in the case of the Internet of Things means that the technical dimension can also facilitate or constrain the enactment of values and the conditions for human flourishing.

Conclusion

Data and the network architectures are far more than mere technical resources. In this blog post I have shown some ways in which the properties and affordances of these technologies that make up the IoT introduce specific challenges and opportunities for developers with clear ethical elements. Looking at cases such as Provenance and Dowse, I have argued that the outcome of the interaction between human values and technical properties is not determined but instead open for innovation and creativity. Further research might want to describe in full detail how this process works. By now, it is possible to say that putting ‘technology’ before ‘values’ in the analysis can bring about a less deterministic and more flexible way of understanding the ethics of the IoT.


References
Bijker, W. E., Hughes, T. P., & Pinch, T. (Eds.). (2012). The Social Construction of Technological systems: New Directions in the Sociology and History of Technology(Anniversary Edition). Cambridge, Massachusetts: MIT Press.
Marres, N. (2012). Material Participation: Technology, the Environment and Everyday Publics. New York: Palgrave Macmillan.
Star, S. L., & Bowker, G. C. (2002). How to infrastructure? In L. A. Lievrouw & S. Livingstone (Eds.), Handbook of New Media: Social Shaping and Consequences of ICTs(pp. 151–162). London: Sage.
Swierstpra, T., & Rip, A. (2007). Nano-ethics as NEST-ethics: Patterns of moral argumentation about new and emerging science and technology. NanoEthics, 1(1), 3–20.