Two emerging areas of computer science – ‘Green IT’ and ‘Sustainable Computing’ – raise some fascinating directions for thinking differently about digital technology and environmental futures. These are attempts to look to nature as a means of reimagining computing – rooted in radically different starting-points from conventional IT industry obsessions with optimising processing power, capability and ‘performance’ at all costs. Instead, ‘Green IT’ and ‘Sustainable Computing’ anticipate out-doors, earthy forms of computing that are enmeshed with ecosystems of water, mud, trees, plants, sunlight and wind. These ideas might sound far-fetched, but are already proving possible. Here, then, are a few different directions that might be pursued in the name of digital degrowth.

i. Powered by nature

One innovative area of sustainable computing is ‘energy harvesting’ – also referred to as energy scavenging and ambient power. This involves computing devices collecting small amounts of energy from background ‘ambient’ sources and therefore set free from the need for batteries or other conventional power sources. This is not a new concept – for example, ‘crystal radio’ sets have been powered by background electromagnetic radiation since the 1910s. Similarly, handheld calculators and watches powered by tiny solar panels became popular consumer products during the 1970s and 1980s. Now these techniques are extended to include computing devices that are powered by micro-wind turbines, small thermoelectric generators, and even devices that collect energy from stray radio waves. Another area of development is computing devices that are powered from kinetic energy generated through people walking, vibration energy harvesting, and even electrically-charged humidity. All these developments follow the logic that computing devices being used in environments that are already full of naturally-occurring energy should not require additional artificially-produced energy.

Alongside harnessing ambient energy, are efforts to power computing directly from natural materials. Perhaps most advanced are efforts to harness computing energy from soil, wetlands and wastewater. Key here is the use of microbial fuel cells -electrochemical cells that gather miniscule amounts of energy from bacteria (exoelectrogens) living in soil and wastewater These microbial fuel cells operate at very low voltages, yet researchers have found this sufficient enough to support the latest forms of low-power computational systems. Interest is therefore growing around the potential of soil-based microbial fuel cells – also known as mud batteries. As Josephson et al. (2022, p.14) put it: “the future of clean computing may be dirty … Practical, large-scale, decades-long deployment of soil powered sensing systems is on the horizon”. 

One of the main promises around powering low-energy computing devices in these ways is ‘untethering’ computing from a reliance on batteries and mains electricity. Again, this is not a new idea – most notably harking back to 1990s’ development of hand-crack and wind-up laptops. Crucially, these principles have real-world potential. A few years back, as a crude proof-of-concept, Dutch engineers developed a ‘battery-free’ Nintendo Game Boy – powered by a combination of sunlight and kinetic energy generated from the console’s buttons (De Winkel et al. 2020). The same team has also developed battery-free Blue Tooth devices that run on harvested energy (De Winkel et al. 2022). 

However, where sustainable computing really gets interesting is applying these principles to support a radically different computing infrastructure beyond a world of Gameboys and other personal devices. Here, we might imagine a sustainable version of the ‘Internet of Things’ with millions of small unobtrusive low-power devices taking sensor measurements, performing calculations and communicating with each other. These devices would be built around ultra-low-power processors that run on harvested energy. Potential applications might include forest-wide networks of sensors that can track the movements of wildlife, or ground-based sensors that can provide farmers with information on growing conditions. Alternately, we might imagine buried and bricked-in sensors that can monitor urban infrastructure and buildings, and even medical sensors that can be implanted and ingested into the human body.

ii. Intermittent computing

These ‘background’ forms of computing that would be capable of looking after themselves constitutes an obvious move away from conventional computer science thinking. In particular, this envisages forms of resilient computing and maintenance-free functioning that is at odds with current forms of high-maintenance and ‘always on’ digital technologies. Indeed, it is incredibly challenging to develop software for devices that are not constantly powered and ‘always on’. This is being addressed through another key area of sustainable computing innovation – what is termed ‘intermittent computing’. Intermittent computing involves the development of devices that slowly harvest and buffer energy when it becomes available, and operate only when enough energy has been banked. Given that ambient sources of energy such as wind, sun and humidity are not constantly available, software needs to be designed to run until the stored energy is exhausted and the device abruptly powers off, waiting for further energy to be harvested (Lucia et al. 2017).

This way of operating throws up some interesting software development challenges. Traditionally, computers are programmed on the assumption that programs will run until they are completed. In contrast, a computer program that is executed intermittently must be capable of spanning power failures – continuously breaking-down and restarting hundreds of times each second. This requires new ultra-resilient approaches to programming that are based around assumptions of inconsistency and imminent breakdown. As Przemyslaw Pawelczak – professor at the Embedded and Networked Systems Group at TU Delft – puts it:

“To allow near-permanent sensing at low cost and at a reduced ecological impact, we have to rethink how we design these systems. We have to let go of the concept of continuous operation. Batteries must be left behind”.

iii. Biodegradable computing

Of course, a world full of digital devices running on ambient energy does little to address the environmental harms related to the production and disposal of circuit boards, electronic components, silicon chips and so on. Another important area of sustainable computing innovation is ‘biodegradable computing’ – organic electronics that are environmentally friendly, low-cost and lightweight (Irimia-Vladu 2014). Again, these are not new ambitions. For example, the development of organic thin-film transistors stretches back to the 1980s, with such materials now commonly found in outdoor displays and lighting, solar cells, e-book screens and smart cards. Now, such principles are being extended into radically different forms of organic computing designed to help us move on from plastics, rare minerals, metals, and other environmentally harmful forms of electronics manufacturing. This includes the development of biodegradable printed circuit boards with recoverable and reusable chips (Arroyos et al. 2022), organic electrochemical transistors, and even the world’s first wooden transistor – produced at the beginning of the 2020s by Swedish scientists from balsa wood (Tran et al 2023). Other materials also being experimented with include biodegradable electronics fabricated from paper, synthetic polymers and silk.

All these forms of technology are far less difficult to manufacture than current approaches. They also incur far less energy and resources in their production. However, adopting these materials as the basis for a new generation of computing forces a rethink of some of the fundamental premises of contemporary mass computing – not least computing speed, processing power and consistency. These organic semiconductors are substantially slower than conventional electronics, with very low electron mobility in comparison to silicon electronics. Biodegradable computing is also far less uniform in terms of performance. Whereas conventional electronics can be produced in a fairly regular manner, individual components that are manufactured from organic materials tend to vary considerably in performance from device to device. 

As such, it makes little sense to compare these new forms of sustainable computing with the mass-produced forms of digital technology that we are currently familiar with. Instead, these biodegradable and organic approaches herald a completely different genre of low-power and ambient computing, with devices embedded in natural environments where they can decompose at the end of their lifespan. This is not a mode of computing where people walk around with multiple personal devices that operate solely for their own use. This is a world where computing is rooted in the natural world, working in the background in the interests of the wider ecosystem. This is not computing that is operated or owned by individual consumers. Put in these terms, then, it should make little difference to anyone whether these background forms of computing are powered by wastewater or mud.

iv. Fungal computing and other forms of ‘wetware’

Perhaps most radical of all are efforts by some sustainable computing researchers to combine low-power computer hardware and software with so-called ‘wet-ware’ – i.e. chemicals, liquids and even biological living systems. Examples here include exotic-sounding ‘slime-mould’ computers, plant computers and chemical computers. Efforts have even been made to develop forms of ‘liquid marble’ computing based on collisions between microliter droplets of liquid. All these different developments share the aim of harnessing the potential of natural materials to structure and support computational processes.

Perhaps most well-known of these are ongoing efforts to harness the electrical activity of mushrooms (more accurately, fungal mycelium) to create computing circuits. Mycelium is the complex web-like root structure of fungi that receive and send electric signals as the fungus grows, as well as retaining memory of these signals. As such, these roots have the potential to act as the electronic components of a computer. Of course, the complex dynamics and system architectures of fungi are radically different from the circuits and wires of conventional computers. This raises the prospect of completely different networking structures that might allow information to be processed and analysed in new ways, and therefore support new forms of computing and sensing. In reality, any form of fungal computing will operate on a much slower basis than today’s machines. However, while lacking in conventional expectations of speed, this is a form of computing that has numerous other advantages – not least the ability to naturally grow and expand, to self-repair and self-regenerate, and only need to consume very small amounts of energy. 

Again, cutting-edge developments such as these suggest radically different approaches to what computing might be. Fungal computing is not intended as a direct replacement for digital computing as we currently know it. Indeed, some computer scientists are working to see how these fungal logics of processing information might be used to improve our current digital information systems. Yet other groups of researchers are working hard to develop fungi computing as a new form of environmental sensing in its own right. Here, it is envisaged that connecting to fungal networks might allow large flows of data to be monitored over wide geographic areas. This could be used to monitor what is going on in an ecosystem – using natural materials such as mycelium as a large-scale underground environmental sensor (Dehshibi & Adamatzky 2021).

Anticipating a new wave of nature-based computing?

All these forms of nature-based computing innovations will undoubtedly seem strange in comparison to the forms of personal computing and digital devices that have dominated over the past few decades. Indeed, one of the leading sustainable computing R&D centres in Mexico goes under the banner of ‘The Unconventional Computing Lab’. Working to develop fungal networks and mud batteries might seem far-fetched to many lay-people, but is no less far-fetched than the idea of cloud-based computing would have sounded if explained to someone in the 1980s. 

Indeed, being open to the idea of new forms of computing built around natural resources like mud, wastewater, balsa wood and fungi should not be too difficult a stretch. After all, our smartphones, laptops and servers are currently dependent on battery packs that are packed full of graphite, lithium salts, metal oxides and other fast-depleting materials. So, why should we not consider much more plentiful resources such as mud and wastewater? Similarly, the idea of natural materials playing a key role in our digital infrastructures is already a reality. Given that our current mode of computing is dependent on networks constructed from copper, then why not be open to the idea of networks made of other materials – especially if they are naturally-occurring and self-sustaining? Mud batteries and plant computing might appear strange, but they are not radically different to the basic material foundations of digital computing that we have become accustomed to over the past 40 years or so. 

All told, these alternate computing innovations open up possibilities of a radically different scale of computing that is in step with our current environmental needs. For example, sustainable computing encourages us to imagine forms of computing that are entwined with nature and integrated into our natural environments. These are forms of computing that do not drain and deplete finite resources. These are forms of computing that are deliberately slowed-down and sporadic, rather than being always-on and hyper-accelerated. These are forms of computing that are designed to decay back into the land and literally become part of the environment.

Crucially, these forms of computing have clear potential in helping care for natural ecosystems and environments. For example, in a world of water shortages and drought it makes good sense to have soil moisture-sensing systems that advise farmers when and where to water their crops. In a world where essential man-made infrastructures face constant risks from extreme weather and natural disasters it seems prudent to have low-powered underground infrastructure monitoring. In a world where we need to preserve as much of the planet’s natural habits as possible, it seems obvious to develop sensing for conservation – capable of anticipating everything from the start of wildfires to supporting the safe movement of wildlife. These are all forms of computing that can work for the good of our ecosystems, not to their detriment. In this age of climate collapse and ecological instability, why would we not want to take this next generation of sustainable computing seriously?

References

Arroyos, Vicente, Maria LK Viitaniemi, Nicholas Keehn, Vaidehi Oruganti, Winston Saunders, Karin Strauss, Vikram Iyer, and Bichlien H. Nguyen. (2022). “A tale of two mice: Sustainable electronics design and prototyping.” In CHI Conference on Human Factors in Computing Systems Extended Abstracts, pp. 1-10..

Chang, Ting-Jung, Zhuozhi Yao, Paul J. Jackson, Barry P. Rand, and David Wentzlaff. (2017) “Architectural tradeoffs for biodegradable computing.” In Proceedings of the 50th Annual IEEE/ACM International Symposium on Microarchitecture, pp. 706-717. 2017.

De Winkel, J., Kortbeek, V., Hester, J., & Pawełczak, P. (2020). Battery-free game boy. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, 4(3), 1-34.

De Winkel, J., Tang, H. & Pawełczak, P. (2022) in MobiSys 2022 – Proceedings of the 2022 20th Annual International Conference on Mobile Systems, Applications and Services. Association for Computing Machinery (ACM), p. 287-301

Dehshibi, M. and Adamatzky, A. (2021). Electrical activity of fungi. Biosystems, https://doi.org/10.1016/j.biosystems.2021.104373

Josephson, Colleen, Weitao Shuai, Gabriel Marcano, Pat Pannuto, Josiah Hester, and George Wells. (2022). “The Future of Clean Computing May Be Dirty.” GetMobile: Mobile Computing and Communications 26(3): 9-15.

Lucia, B., Balaji, V., Colin, A., Maeng, K., & Ruppel, E. (2017). Intermittent computing: Challenges and opportunities. 2nd Summit on Advances in Programming Languages (SNAPL 2017).

Mihai Irimia-Vladu. (2014). Green electronics: biodegradable and biocompatible materials and devices for sustainable future. Chemical Society Reviews 43(2):588–610. 

Tran, Van Chinh, Gabriella G. Mastantuoni, Marzieh Zabihipour, Lengwan Li, Lars Berglund, Magnus Berggren, Qi Zhou, and Isak Engquist (2023) “Electrical current modulation in wood electrochemical transistor.” Proceedings of the National Academy of Sciences 120(18): e2218380120.