Site icon Conservation news

Beyond data collection — the social and political effects of environmental sensor proliferation

  • Sensors are increasingly collecting data everywhere, changing how we relate to and manage the environment.
  • These data answer researchers’ scientific questions but can also generate social, cultural and political effects, reinforcing the need for more data.
  • In a new book, “Program Earth: Environmental Sensing Technology and the Making of a Computational Planet,” Dr. Jennifer Gabrys suggests that the future of environmental management is a question of how and whether to “program” the Earth.

Look across the pages of this website and others which feature emerging technologies, and you find a world that is increasingly “wired up” with interoperable, multifunctional devices that “sense” the environment, providing scientists with much needed data. Sensors you already know include cameras, GPSs, smart phones, thermometers, webcams and acoustic monitors.

As these and other technologies become more affordable and widely dispersed, greater numbers of citizen scientists are using sensor apps and tools to collect information about their environment, effectively crowdsourcing the building of knowledge about the world.

Compiling these data produces a picture of the status of endangered species, climate, pollution and environments of concern within a sensor network, creating a “smart” environment, as wired-up devices can respond to changing external conditions by making automatic adjustments according to pre-programed algorithms and software programs.

Where is all this technology going? Are we really headed towards a version of Earth that functions like or even according to a computer program?

To learn more about how increasingly ubiquitous sensing is informing and affecting our relationship to the environment while making possible the prospect of a programmable Earth, WildTech spoke with Dr. Jennifer Gabrys, whose recent book Program Earth: Environmental Sensing Technology and the Making of a Computational Planet examined the implications of this proliferation of environmental sensors. Gabrys, of the Department of Sociology at Goldsmiths, University of London, is also Principal Investigator on the European Research Council (ERC) project, “Citizen Sense,” which investigates the use of environmental sensors for new modes of citizen involvement in environmental issues.

Dr. Jennifer Gabrys. Photo credit: Holly Revell

WildtechHow do sensors affect the way we relate to one another and to the environment?

Gabrys:  In Program Earth, I discuss the ways in which sensors organize or inform experiences. Rather than see sensors as technologies that are merely describing a reality “out there,” I suggest that sensors “tune” our attention to environments and environmental problems in particular ways. This is not necessarily to say that sensors are generating fictitious accounts of environments, but rather that sensors structure our approaches to environments and environmental problems.

Whether we are attempting to address climate change or deforestation, we often encounter these problems through the perceived need to gather more data, which sensors (both on-the-ground and remote) are meant to facilitate. This way of organizing scientific inquiry also affects the ways we relate to environmental problems as problems of generating particular types of evidence.

WildtechWhy do we need sociologists and humanities scholars to help us think about wildlife, pollution, and urban environmental sensors? Aren’t these just straight-ahead data collection devices?

Inevitably, problems of environmental data and data collection are more than just scientific topics.

Scholars in the social sciences and humanities have been particularly insightful in addressing how casting something as a problem of data collection can be a way to depoliticize environmental problems through the apparently neutral modes of scientific practice. Scholars in these fields also have demonstrated how data collection is bound up with decisions, values, evidential claims and politics, since data is also about power and who has the ability to make claims. This is very evident in citizen science projects that collect data and attempt to challenge official accounts.

WildtechYou write, “Sensors are everywhere. Small, flexible, economical and computationally powerful, they operate ubiquitously in environments.” Could one argue that we already have too many sensors?

Rather than ask whether there are too many sensors, I would instead wonder why sensors are now proliferating and being used in particular ways. While I’ve noted the ways in which environmental problems have come to be seen as problems of data, there are also many ways in which sensors are used to actuate particular responses and manage environments for greater speed and efficiency. This is the case in smart city scenarios.

So we might ask, what is the work that sensors are doing? What is gained and what is lost when thinking about environmental problems in terms of more data to be gathered? And how does the sensorization of environments create other problems, whether to do with surveillance, hackability of infrastructures or the resource requirements of electronic systems?

WildtechWhat role do sensors play in utopian visions of sustainability?

Sensors are often bundled into a certain kind of sales pitch for the Internet of Things, where smart environments are meant to allow us to achieve a more balanced functioning of environmental systems. From smart cities to smart meters, sensors operate as part of data-managed infrastructures that are meant to deal with resource shortages, population pressures and the perceived need for greater efficiency.

Here, again, we might say that a technological solution is imagined for what are sociopolitical and environmental issues. I write about some of the drawbacks of this technological approach to sustainability in my discussion of smart cities in Program Earth. However, at the same time there are many ways in which sensorized environments fail to achieve the desired outcomes, whether because efficiencies can actually drive new forms of consumption in other areas or because technologies fail, require updates, become obsolete and have considerable resource requirements of their own from manufacturing to use and disposal. This is a topic I addressed previously through a book on electronic waste, Digital Rubbish: A Natural History of Electronics.

Technology challenges: a camera trap destroyed by elephants in Pakke Tiger Reserve, India. Photo credit: Nandini Velho CC BY-SA 3.0

WildtechDo you see the overall movement towards a Program Earth a good or bad thing? What should we keep in mind about programming and about the Earth when thinking about this question?

The title, Program Earth, comes from the suggestion that when satellites were launched, a particular view of the Earth as a programming problem came into being. When the Earth is sensed in particular ways, it also becomes configured for a particular set of actions and relations. I look at some of the ways of “programming” the Earth to consider what these approaches mean for our understandings of environments and environmental problems, whether through tracking migrating animals, tracking ocean currents and marine plastics or monitoring the effects of climate change.

WildtechWhat surprised you most in your fieldwork for the book?

Perhaps the sensor testbeds at James Reserve in California that were installed as part of the NSF-funded Center for Embedded Networked Sensing (CENS) project. I had done a considerable amount of reading about the sensors before arriving at the site and visited researchers at their lab at UCLA, and so I had a developed a vision of the sensors as blanketing the landscape and robots crossing waterways and tree canopies. Some of the early diagrams—one of which is in Program Earth—show the sensors operating in just that way.

When I arrived at the ecological reserve, however, I found that the sensors were not as pervasive as this and were clustered in particular study sites to analyze soil moisture or respiration or gather data on phenology in relation to a few selected organisms. The sensors were deployed selectively in relation to research questions, which shows how there is a tuning across sensors, environments, research and researchers in order to home in on particular environmental problems.

A water quality sampling station installed along the East Branch Milwaukee River in Wisconsin, USA. The station collected samples either at programmed time intervals or proportionate to flow over a specified period. The data logger (white cabinet) recorded temperature, specific conductance, and dissolved oxygen. Photo credit: USGS

WildtechCan you discuss some of your favorite smartphone apps from the book? What were some of their notable intended and unintended uses and effects?

There are now an incredible number of apps focused on ecology, citizen science and environmental awareness. I focused on some of the apps more aligned with environmental research. One that I found particularly interesting was Max Planck Institute for Ornithology’s AnimalTracker, which allows users to follow tagged migrating animals—in this case, storks—to see how their movements shift in relation to seasons and how these movements intersect with different land use issues in different areas.

Some of the scientists on this project have proposed that they would like to see “every animal with a cell phone,” or in other words, that animals would be tagged and tracked for their entire lifecycles. I wondered whether the drive to collect this data could make a user witness to the possible decline and disappearance of organisms without necessarily generating a way to act on that data or improve the environmental conditions of these organisms.

WildtechYou spend a chapter discussing two particular webcams: one of moss in a reserve and one of the BP spillcam. Can you explain why these are worthy of your focus and reflection? What work, besides the obvious recording, are the sensors doing?

One of the things that I found fascinating while researching the James Reserve and the CENS project was how cameras were used as sensors to study environmental change and phenology. Even traffic cameras became newly interesting technologies since the image data available from these could be used to study changing plant life, for instance, in the view of these cameras across entire geographical areas.

I look at two webcams, the Moss Cam and the Spillcam, in order to consider the variety of sensor data that emerge with webcams. With the Moss Cam at the James Reserve, a web camera fixed on a moss specimen was both sensing changes in this bryophyte and part of a larger sensor ecosystem where infrared cameras, temperature and humidity sensors as well as other web cameras, such as bird cams, could be linked up to begin to make inferences about phenology or even to discover new ecological events such as the surprising relationship of a chipmunk to this particular patch of moss.

The Moss Cam at webcam and infrared camera, which analyzes the health and seasonal condition of the moss at James Reserve. Photo credit: Jennifer Gabrys

While the Spillcams that were set up as a response to the BP oil spill were on one level operating as sensors that tracked the rate of flow from the ruptured wellhead, on another level they were also tools of political accountability. The Spillcams were meant to provide evidence so that BP could be held to account but also became devices for witnessing a disaster that people felt powerless to address.

WildtechI was struck by your provocative question, “Can a fish be a citizen?” Do sensors turn animal study subjects into citizens?

In posing this inevitably provocative question, I am deliberately trying to ask how we configure ourselves as political subjects and as environmental citizens. The default way of thinking about citizenship is as a unit of belonging to a nation-state, where certain political engagements are possible and expressed. But if we open up our thinking about citizenship to consider how new political subjects might form through ecological relations, we could at once broaden the scope of our belonging and accountability to environmental affairs while recognizing that other entities are expressing the effects of environmental change in particular ways and in relation to their particular inhabitations.

Citizenship in this sense is not only a matter of being a singular political subject, but also of belonging to extended environmental communities. For instance, if organisms such as lichens are bioindicating the effects of air pollution, they will be [reflecting] their own lifecycles, the impact of pollution on forests or other landscapes, and its effects on humans.

Setting up a wildlife camera trap in the rainforest. Photo credit: G. Powell

WildtechAre humans necessary in Program Earth?

This question is almost as provocative as asking if fish can be citizens! One could certainly imagine a science fiction scenario where sensor-actuators determine that the most efficient and sustainable environments would be ones where humans would be managed in particular ways—perhaps even to the point of their elimination. And there are no shortage of AI scenarios that are wondering and worrying about this very possibility.

As I suggest in Program Earth, drawing on the work of French engineer and philosopher Gilbert Simondon, we might instead consider how a programmed earth causes us to rethink and rework what counts as “human.” Simondon’s point is that the human is not an automatically settled category, but is something that emerges in relation to environments, technologies and other entities. We might then consider how we could find and practice new articulations of the human that are more aligned with environmental survival for ourselves and other organisms.

WildtechWhat should we consider and/or be concerned about in this period of expanding technological innovation?

Sensors and the Internet of Things are now developing apace, both in environmental science and numerous other environmental applications. Beyond the list of issues mentioned above, there is also the problem of the surveillance of humans and nonhumans—in fact, some scientists are not publishing the results of their tracking studies of nonhumans, due to concerns about poaching and wildlife trafficking—as well as the increasing vulnerability of sensorized infrastructures to hacking, as the recent Mirai botnet attack demonstrates.

While more data and more sensors do not automatically lead to more insights, despite the tech company straplines, they do give rise to a potentially renewed engagement with the intersections of environment, technology and politics. It is important that more people take up these topics of sensorized environments and work in a transdisciplinary way to open up these technologies to other engagements. I hope this is what we are realizing in part through the Citizen Sense project, where we are asking how a more democratic engagement with these technologies can present new ways to think about and express environmental citizenship.