In this post I dissect the idea of sustainable safety and security, starting from the questions of what we want to sustain, and for whom we want to sustain it. I believe that a general "people and planet" answer is inadequate here because this form of sustainability cannot be the property of a single industry sector but must be addressed by society as a whole. However, with sufficient understanding of life-cycle impacts we may very well be able to devise research and development efforts, and inform decision making processes towards the use of integrated safety and security solutions that help us to address societal challenges in the context of the climate and ecological crises, and that are aligned with concepts such as intersectionality and climate justice. Of course, these solutions can only be effective if they are embedded in societal and economic change towards more frugal uses of data and ICT.

Abstract

Security and safety of system are important and intertwined concepts in the world of computing. Where security aims to protect systems from intentional attacks, safety strives to ensure systems’ operation under unintentional hazards. With the increasing societal dependence on digital technologies and the growing threat that changing environmental conditions and cyber attacks pose to our infrastructures, we understand that safety and security are linked concepts that critical systems but also consumer electronics must implement.
In recent years, the terms “sustainable security” and “sustainable safety” came into fashion and are being used referring to a variety of systems properties ranging from efficiency to profitability, and sometimes meaning that a product or service is good for people and planet. This leads to confusing perceptions of products where customers might expect a sustainable product to be developed without child labour, while the producer uses the term to signify that their new product uses marginally less power than the previous generation of that products. Even in research on sustainably safe and secure ICT, these different notions of terminology are prevalent. As researchers we often work towards optimising our subject of study towards one specific sustainability metric – let’s say energy consumption – while being blissfully unaware of, e.g., social impacts, life-cycle impacts, or rebound effects of such optimisations.
In this paper I dissect the idea of sustainable safety and security, starting from the questions of what we want to sustain, and for whom we want to sustain it. I believe that a general “people and planet” answer is inadequate here because this form of sustainability cannot be the property of a single industry sector but must be addressed by society as a whole. However, with sufficient understanding of life-cycle impacts we may very well be able to devise research and development efforts, and inform decision making processes towards the use of integrated safety and security solutions that help us to address societal challenges in the context of the climate and ecological crises, and that are aligned with concepts such as intersectionality and climate justice. Of course, these solutions can only be effective if they are embedded in societal and economic change towards more frugal uses of data and ICT.

A PDF version of this post is available as

The text is partially based on a keynote talk I gave in March 2022 at the Workshop on Sustainability in Security, Security for Sustainability (SuSec), co-located with DATE 2022:

Introduction

Our total reliance on computing infrastructures becomes most apparent when those infrastructures fail. I am writing these lines while waiting at a train station in The Netherlands, on Sunday the of April 2022. There is free coffee, but pretty much no trains are operational. NS, the main network operator, announced earlier that due to a technical problem there will be no trains running until 17:00. No further details are revealed at this moment but speculations are going wild: “It’s probably an attack,” someone says. Across the country, thousands of people are stranded at stations. Alternative means of public transport cannot cope with the rush demand, roads are jammed with taxis and personal vehicles. “Due to the enormous impact of the failure in the IT system, it is unfortunately not possible to run any trains today,” we hear later. The failure “affected the system that generates up-to-date schedules for trains and staff. This system is important for safe and scheduled operations: if there is an incident somewhere, the system adjusts itself accordingly. This was not possible due to the failure.”1 At least we had free coffee.

"Sunday 3 April no more NS trains." Source: ns.nl
"Sunday 3 April no more NS trains." Source: ns.nl

As our society is becoming increasingly dependent on digital technologies, research and industry made important efforts to ensure the safety and security of these technologies. Here safety is the property of a system to achieve an acceptable level of risk by controlling recognised hazards – predominantly environmental conditions or human error, which are expected to behave within certain parameters. In the context of Information and Communication Technology (ICT), safety thus involves properties of the hardware (e.g., ensuring safe operating temperatures for a processor) and software (e.g., ensure correct execution and safe behaviour of the software scheduling NS trains and staff). Security is distinguished from safety and denotes the property of a system to withstand intentional attacks from an intelligent adversary. Such an adversary will seek to interact with the system in unexpected ways to make the system misbehave, often violating safety as a result.

The fundamental distinction between security and safety is regarding the presence (or absence) of an intelligent and purposefully malicious adversary, their ability to act outside of the limits of safe interactions considered when the system was designed. So, when the Ukrainian power grid was hacked in 2015, the attackers “were skilled and stealthy strategists who carefully planned their assault over many months, first doing reconnaissance to study the networks and siphon operator credentials, then launching a synchronised assault in a well-choreographed dance,”2 which left 230,000 residents without electricity.

Security and Safety as Emergent Properties

With this background it becomes clear that security and safety are emergent properties that only apply to a system that is being used in a specific context and environment: A processor may safely operate under the expected outdoor temperatures in Belgium but not in Ethiopia. A control software may be acceptably secure to operate on the isolated local network of a factory but not over a public network where an attacker may modify or replay communication. It also becomes clear that safety and security are heavily dependent on assumptions that may be violated once in a while, e.g., by increasingly high summer temperatures in Belgium, or by an attacker who unexpectedly gains access to a seemingly isolated network. As such, neither safety nor security are absolute but are about reducing risk levels to be acceptable by society and within the margins of standards and regulations.

Furthermore we see that safety and security are interconnected, in particular, that safe operation of equipment can only be guaranteed if that equipment is also secure. The field of study that combines these two aspects of risk assessment and engineering is dependability engineering (Council & others, 2007), and important progress has been made regarding safety and security co-assurance for critical systems (Gleirscher et al., 2022). Particularly hard to assess in this context is the long-term survivability of systems, as Eder-Neuhauser et al. point out with respect to the security of power grid infrastructure in (Eder-Neuhauser, Zseby, Fabini, & Vormayr, 2017): “However, unlike consumer electronics, traditional power grid environments have a focus on long-term stability and plan for hardware life-spans of 10 years or more. As devices age, unknown vulnerabilities of hardware, operating system, software, and protocols emerge. Such vulnerabilities pose a serious threat to the infrastructure. While consumer electronics need not fulfil the same life-cycle requirements of industrial devices, their base technology is similar.”

This Blog Post

In this paper I highlight and discuss the relevance of security and safety for sustainability in the ICT sector. Recently, more researchers and companies strive to link their output – be it research, equipment or services – to varying definitions of sustainability, based on characteristics and measurements that involve efficiency, overheads, profitability, or environmental impacts. Reflecting on these characteristics, I analyse the idea of sustainable safety and security, starting from the questions of what we want to sustain, and for whom we want to sustain it. I hope that this discussion can contribute towards developing an integrated understanding of safety and security that helps us to address sustainability topics and societal challenges in the context of the climate and ecological crises. I believe that this understanding may then allow us to align technical development with concepts such as intersectionality and climate justice. What becomes clear is that sustainability efforts that target security- and safety engineering can only be effective if they are embedded in societal and economic change towards more frugal uses of data and ICT.

Notions of Sustainability in the Context of Security and Safety

In (Paverd et al., 2019), Paverd et al. link security and safety, and implicitly dependability, to a notion of sustainability that focuses on maintaining key properties of systems by preparing them for threats not known today, defining design principles for a vision of sustainable security and safety. These design principles focus on diversification, replicability and relocatability of resources, containment of failures, adaptability and updatability, with simplicity, verifiability and minimisation of assumptions as overarching concepts. While there are many open research challenges regarding the implementation of these principles, following the vision of Paverd et al. would lead to the development of software and equipment designed for safe and secure longevity.

On Costs and Benefits

System designs resulting from following the sustainable security and safety principles from Paverd et al. (Paverd et al., 2019) would likely incur increased costs regarding the development, initial equipment purchase, and the maintenance of installations. For example, modern processors typically come with built-in support for encryption. Since about 2010, processor vendors implement instructions for the AES cipher (Daemen & Rijmen, 1999) in hardware, and ever since viable attacks against these hardware implementations are being published. These attacks necessitate the replacement of equipment where the risk of such attacks is deemed unacceptable. A way to potentially limit hardware replacement is by using updatable hardware and crypto agility, which may come with different security guarantees and more complex equipment in the first place.

Another well-known example for such issues are the Spectre and Meltdown vulnerabilities (Kocher et al., 2018; Lipp et al., 2018), which render entire generations of processors vulnerable to software-level attacks. Processor vendors and operating system developers have provided patches to mitigate these vulnerabilities at the expense of a substantial degradation of performance and increase of energy consumption (Herzog, Reif, Preis, Schröder-Preikschat, & Hönig, 2021; Prout et al., 2018; Alhubaiti & El-Alfy, 2019). With application-specific energy overheads of up to 72% (Herzog, Reif, Preis, Schröder-Preikschat, & Hönig, 2021), applying these patches in existing installations raises questions of resource availability and scalability, and may necessitate a lot of additional equipment to be added to compensate for the loss of processing power. Indeed, systems that are designed by following the sustainable security and safety principles from (Paverd et al., 2019), in particular diversifiability, relocatability, and containment of breaches, can very likely be adapted and kept secure under this changing attacker model, but this maintenance will still incur substantial costs.

We realise that, with layers and layers of patches and security features built on top of each other, and that all require continuous maintenance and re-evaluation to maintain reasonable levels of security for internet-connected systems, these costs build up quickly and contribute to the overall short lifespan of ICT equipment.

Sustainability as a Requirement

Sustainability as a first-class quality of ICTs, on an equal footing with safety and security, has first been proposed by Penzenstadler et al. in (Penzenstadler, Raturi, Richardson, & Tomlinson, 2014), arguing that “instead of merely optimising current systems, software engineers must embrace transition engineering – an emerging discipline that enables change from existing unsustainable systems to more sustainable ones by adapting and filtering demand to a declining supply.” Here, transition engineering is a discipline that aims at “identifying unsustainable aspects of current systems, assessing the risks posed by those aspects, and researching and developing ways to mitigate and prevent systemic failures through adaptations” (Krumdieck, 2011). Penzenstadler et al. point to a conflict between organisational goals on the one side and security, safety, and sustainability on the other, because these qualities are commonly perceived as barriers to profitability; predominantly so because companies do not pay for environmental and social impacts of their products. In this context, sustainability can be seen as a non-functional requirement similar to security or safety, the implementation of which requires the establishment of a sustainability culture in organisations and in society.

There is a growing body of literature on requirements engineering for software systems that engage with the multi-dimensional nature of sustainability, e.g. (Becker et al., 2016; Venters et al., 2018; Duboc et al., 2019). Albeit with no specific considerations for safety and security, it is noted that systems resulting from such efforts are “different when sustainability principles and therefore long-term consequences are considered” (Becker et al., 2016). As safety and security are emergent properties of systems that involve qualities of, and interactions between, a system’s hardware, a system’s software, and the operating environment, engineering processes that aim at sustainably safe and secure systems need to involve approaches to hardware-and-software co-design based on requirements catalogues such as Paverd et al. (Paverd et al., 2019). Notions of sustainability and circularity in micro electronics that consider the physical basis of ICT infrastructures have, e.g., been developed by Griese et al. in (Griese, Schischke, Reichl, & Stobbe, 2004), and Clemm et al. in (Clemm et al., 2019), which need to be integrated with the work on sustainability for software systems mentioned above.

Earlier notions of sustainability in ICT lead back to Köhler and Erdmann (Köhler & Erdmann, 2004) and Hilty et al. (Hilty et al., 2006), who distinguish three dimensions where software systems impact sustainability: (1) effects of the physical existence of ICT, such as the impacts of the production, use, recycling and disposal of equipment; (2) indirect environmental effects of due to changed processes in other sectors, including e.g., the optimisation of industrial processes or uses of products and services which in turn result in changed environmental impacts of these processes, products or services; and (3) tertiary effects resulting from the medium- or long-term adaptation of behaviour such as consumption patterns or business models, as a result of the continuous availability of ICT products and services. In summary, Hilty et al. (Hilty et al., 2006) conclude that “the overall impact of ICT on most environmental indicators seems to be weak, the impact of specific areas or types of ICT application can be very relevant in either direction. On an aggregated level, positive and negative impacts tend to cancel each other out.” More research is certainly necessary to understand whether these conclusions still hold in the light of current socio-economic and environmental impacts of ICT sector, for example the online advertising sector (Frick, Matthies, Thøgersen, & Santarius, 2021; Cucchietti, Moll, Esteban, Reyes, & García Calatrava, 2022), being an engine of economic growth and a key protagonist to sufficiency-oriented consumption.

The above definitions and aspects of sustainability are all focused on ICTs. The most generally applicable, and certainly the most quoted notion of sustainable development (not sustainability in general) comes from Brundtland (Brundtland, 1987), referring to “development that meets the needs of the present without compromising the ability of future generations to meet their own needs.” In difference to the definitions discussed before, Brundtland puts an emphasis on societal needs. If these are no longer satisfiable, what we understand as society right now, may cease to exit. Thus, maybe we should not see sustainability as a quality of ICT systems, besides security and safety, but use Brundtland’s definition as the overarching safety requirement for all ICT development?

Conflicting Expectations

While all these diverging notions of sustainability and sustainable safety and security make sense with respect to their specific frames of reference, they are surely confusing for everyone who simply wants to buy a sustainable product or use a sustainable service. As suggested by sparsely available empirical research, public understanding of sustainable development may centre around “green issues,” with a focus on saving energy and reducing waste (cf. (Scott, 2015)). Of course, other interpretations are possible and one might expect a product to be manufactured under fair labour conditions or with responsibly mined resources. Now, while a manufacturer might label their business as sustainable on the basis that they are able to continuously increase employee salaries, a consumer might expect the product to be somehow “good for the planet.” To make things worse, anecdotal evidence shows that there is also “secret sustainability,” where companies do not announce their sustainability efforts out of fear that consumers would expect their more sustainable products to be inferior to less sustainable competitors, either in terms of a reduction in product quality, or an increase in the price of manufacturing.3

We cannot barter notions of sustainability for safety and security. Instead we must strive to design and use technologies so as to reduce the divide between the over-served and those who are marginalised by society, economic development, or technology.

Of course, all these notions and considerations are fairly generic. What could it now possibly mean if a product or service features sustainable security and safety? Does it mean that the products jointly manages risks related to safety, security, and sustainability, that it sustains its safety and security properties, or that it implements safety and security in some way that is good for “people and planet?” I want to dissect this question by asking explicitly what it is that we sustain, and for whom we sustain it.4 Understanding that personal safety and security, together with privacy, are fundamental human right listed in the Universal Declaration of Human Rights (UDHR) makes asking these these questions even more important: We cannot barter notions of sustainability for safety and security. Instead we must strive to design and use technologies so as to reduce the divide between the over-served and those who are marginalised by society, economic development, or technology.

Sustaining What? And for Whom?

Predicting the impact of security and safety technologies on the different dimensions of sustainability – society, environment, culture, and economy – is difficult. Thus, it is not surprising that research in security and safety hardly ever bothers with such analyses but focuses on reporting more easily quantifiable evaluation metrics such as security under different attacker models, performance and overheads, system complexity or the price of installations.

But would an analysis beyond these metrics even be useful? Is it not the case that managing security and safety risks in any product will always incur overheads, will always increase the environmental footprint of a product, and will always ask people to change their behaviour? Does sustainable security and safety then make sense as a term or concept, or is it merely a framing that we as security and safety professionals use to align our work with a current marketing trend that aims to make consumers feel good with their sustained consumption? Or should we seek to carbon-offset the impacts of security and safety mitigations and report in our papers the number of trees we planted for every microcontroller purchased for some experiment?

I fundamentally believe that sustainability in the general sense of “meeting the needs of the present without compromising the ability of future generations” is not a property of an individual engineering discipline or industry sector. Even if the ICT sector would reach net-zero emissions, it would still not be sustainable in itself due to its demand for virgin resources and exploitative labour, and the environmental and social impacts related to these.5 We understand that as a society we can reach and maintain Brundtland’s notion of sustainable development by operating within the planetary boundaries (Rockström et al., 2009), by defining our economic activities in the doughnut (Raworth, 2017), and we understand that ICTs can contribute to this: “Infrastructure design and access, and technology access and adoption, including information and communication technologies, influence patterns of demand and ways of providing services, such as mobility, shelter, water, sanitation, and nutrition. Illustrative global low demand scenarios, accounting for regional differences, indicate that more efficient end-use energy conversion can improve services while reducing the need for upstream energy by 45% by 2050 compared to 2020.” (Grubb et al., 2022)

We occupy a position of power that we must use and strive to provide adequate notions of safety and security that are inclusive, that serve society as a whole, and that do not make parts of the world pay for the advancement of others.

The notion of sustainability that relates most closely to ICT systems follows Tainter (Tainter, 2006), who sees “sustainability is an active condition of problem solving, not a passive consequence of consuming less.” Ultimately it is the responsibility of society and our democratic processes to define what resources we are willing to spend on critical infrastructures and for individual consumption, and which role security and safety needs to play in this framework. And if we as society, in such a modelling and planing exercise, and with the necessary democratic legitimacy, conclude that every citizen may procure a new smartphone every ten years (following a model from (Millward-Hopkins, Steinberger, Rao, & Oswald, 2020)), then it must be the task of the ICT sector to design a device that can be operated with adequate security guarantees – defining that term also requires democratic legitimacy – for at least the anticipated duration of use.

However, the reach of ICT is pervasive and few industries have shaped society as much as modern communication systems and automation. This huge societal impact must come with equally huge responsibility, and as researchers, engineers, or companies active in this field, we occupy a position of power that we must use and strive to provide adequate notions of safety and security that are inclusive, that serve society as a whole, and that do not make parts of the world pay for the advancement of others.

Engineering Practice: Cats and Mice

The filed of systems security is particularly remarkable in this regard: Security is generally perceived as a cat-and-mouse game where whatever security measures are being put in place and are (almost always) defeated by a sufficiently advanced and resourceful attacker in the foreseeable future. As such, it is impossible to infinitely sustain the security of some system, nor is is possible to make security sustainable in the “people-and-planet” sense. In (Mühlberg & Van Bulck, 2018) we argue, with respect to processor security in the context of the aforementioned Spectre and Meltdown attacks, that this cat-and-mouse cycle can possibly slowed down by following a more principled approach to designing security solutions, which would likely follow the criteria for “sustainable security and safety” laid out by Paverd et al. (Paverd et al., 2019).

Nevertheless, sustaining security would still be energy- and material intensive. We have to understand that, on the one hand, security mechanisms may serve a very specific demographic, namely those who can afford to care and for whom the economic benefit of expensive security comes with the promise of even greater return on investment. On the other hand, implementing security that is designed for this one specific demographic, may put a potentially much larger group of people at the risk of extended exploitation: for example those people whose digital security is not deemed worthy of the investment, those who work in artisanal mines in the global south, extracting the raw materials for new and more secure electronics without ever being able to afford a device featuring the latest security features, and also those for whom a new security solution may require unreasonable adaptation of their social behaviour. As Wu et al. point out in (Wu, Edwards, & Das, 2022), “ignoring human social behaviours in designing [security and privacy] systems leads to maladaptive user behaviours that either reduce security, cause social friction, or both.”

Interestingly, safety emerged as a very different discipline of engineering. Being focused on mitigating adverse environmental conditions and human error within constrained parameters – users make benign errors but do not actively seek to disrupt the functioning of a system – safety engineering did traditionally not exhibit the characteristics of the aforementioned cat-and-mouse game as much as security does. As such, managing safe operation of a system requires less of a continuous effort: Once the safety requirements for a specific operating environment and risk vectors are established and implemented, the system will only require maintenance when components fail. Under harsh conditions that may of course happen regularly, but it is still predictable and the mitigations to failure – maintenance windows, replacement components, etc. – can be planned in advance. This approach to safety, however is changing. With the increased connectivity and distributedness of safety-critical systems, these systems must focus more and on dependability and integrate security engineering as a prerequisite for system safety.

The “What”

Therefore the key to sustaining security and safety probably is to be very cautious about defining the assets that need protection: What do we want to sustain? The business models of companies? The security of a smart grid infrastructure that enables a community to optimally use renewable energy sources? The confidentiality of amassed profiling data and machine learning models, collected and processed with the purpose of learning the data subjects’ intimate desires to produce targeted advertising and sell them more things to perpetuate economic growth? Threat modelling (Shostack, 2014) is the discipline of security engineering that is meant to give us an understanding of assets, actors, and the required threat mitigations, albeit without a critical reflection on whether these assets need to be there in the first place. What is required on top of this is a notion of critical refusal, a challenging reflection on harmful practices of building infrastructure, data collection and the like, while opening up spaces to develop alternative proposals (cf. (Simpson, 2007; Theilen et al., 2021)). Such notions of refusal have, e.g., been expressed in the Feminist Data Manifest-No (Cifor et al., 2019), which could very well be appropriated by researchers and practitioners in safety and security engineering.

Importantly, alternative approaches in this domain do not exclude new ICT systems or the data-based optimisation of businesses or data per se. Moving towards data collections and inference that serves the public interest, that is being collected and processed following data protection and privacy by-design principles (Gürses, Troncoso, & Diaz, 2011), and that is then being made available under an open data policy, could already lead to enormous reductions in the need for security infrastructure. That is because such an approach would enable sharing and reuse of data and infrastructures, and would reduce the need for confidentiality protection, with only the need for integrity protection and availability remaining.

Reflections towards critical refusal of technical developments and towards reducing the quantity of assets that need protection, essentially enabling us to, e.g., switch data centres off instead of patching thousands of machines and then mitigating the overheads by adding more machines, potentially have a tremendous positive impact on societies digital security requirements and also reduce environmental impacts dramatically, much more than technical optimisations in the implementation of cryptographic algorithms and the like. Also, if protecting the confidentiality of assets is no longer a security objective that need to be upheld for a majority of assets, it may be feasible to operate (parts of) data centres without patching confidentiality-related vulnerabilities such as, in certain scenarios, Meltdown and Spectre.

The “Who”

Technology affects people differently. If we, for example, develop a product that aims to sustain road safety for fast vehicles by introducing a vehicular communication system that allows for coordinated emergency braking (Cao, Gangakhedkar, Ali, Gharba, & Eichinger, 2016), we probably impose additional risks on road users to whom this technology is not available. Pedestrians, cyclists, and users of older vehicles may see themselves endangered by a new generation of faster vehicles whose drivers show little concern for the presence of vulnerable road users. Or we could think of an implementation of onion routing (Goldschlag, Reed, & Syverson, 1999) to sustain users’ privacy when browsing the Internet, which comes with increasing material and energy consumption due to additional layers of cryptography and additional routing hops for network packets. In consequence, this privacy tool may not become equally available to everyone and may never benefit, or indeed impose additional burdens, on the most vulnerable members of society, internationally. Yet, the availability of this tool has been important, e.g., in the context of whistleblower protection and to help those organising and fighting oppressive regimes.

Specifically when looking at the North-South divide regarding the availability of technology we see dramatic differences in who has access to which technologies, whom technological advancement serves, whom we develop technology for, and respectively who controls technology roll-out and platforms. This all affects how we define sustainability in ICT. My institution, for example, collaborates with an initiative to collect and refurbish ICT equipment, typically after three of four years of university use, to give them a second life in the Global South.6 For what looks like a great initiative at a first glance, we have to ask if this practise can serve as a sustainability goal or if it instead indurates current power structures. Why is it that people in the Global South are supposed to be fine with four-year-old equipment while we cannot use that equipment any further? Why is it that those economies, where most of the mining of basic resources for our equipment is done, at horrendous social and environmental costs, cannot afford new equipment but depend on our donations? Can safe and sustainable waste management be guaranteed when the equipment eventually reaches its end of life in the destination country? How is security managed for used equipment where hardware and built-in software or credentials could be compromised? Do such equipment donations actually aim at mitigating power imbalances between the Global North and the Global South?

A Concrete Example

A smart irrigation system as an example for distributed application networks we envisage.
A smart irrigation system as an example for distributed application networks we envisage.

Last year, my colleagues and I published a paper (Alder, Van Bulck, Piessens, & Mühlberg, 2021) in which we develop a security architecture that provides a notion of guaranteed real-time execution for dynamically loaded programs that are isolated in so-called enclaves. These enclaves are are a concept developed in the context of trusted computing – techniques to execute code in isolation from other software, even from the operating system. On top of isolation, trusted execution environments provide primitives to convince a remote party of the integrity of the enclave and whatever computations this enclave might perform (cf. (Maene et al., 2017) for a survey). We implement preemptive multitasking and restricted atomicity on top of these enclaves, illustrating separation of concerns where the hardware is to enforce confidentiality and integrity protections, while a small enclaved scheduler software can enforce availability and guarantee strict deadlines of a bounded number of protected applications. The approach is designed for open systems, where multiple distrusting actors may run software on the same processor, and without introducing a notion of priorities amongst these mutually distrusting applications. We implement a prototype on an extremely light-weight open-source processor, and illustrate in a case study that protected applications can handle interrupts and make progress with deterministic activation latencies, even in the presence of a strong adversary with arbitrary code execution capabilities. We envision our processor designs to be used in the context of heterogeneous control networks that implement distributed sensing and actuation for critical infrastructures, for example in the context of smart farming or to facilitate the transition towards renewable energy sources. In (Scopelliti et al., 2021) we outline such a scenario.7

Our two papers, (Alder, Van Bulck, Piessens, & Mühlberg, 2021) and (Scopelliti et al., 2021), come without a sustainability evaluation, and I have not seen a conference in the security, safety, and dependability domain that asks for such an evaluation to be part of submissions. One reason for this is probably the lack of established criteria to do so. In our case, evaluating our work along the lines of the relevant aspects of the sustainable security and safety principles laid out by Paverd et al. in (Paverd et al., 2019) would be a reasonably thing to do. Alternatively, one could also look at specific use cases and argue how these contribute towards achieving the UN’s Sustainable Development Goals (Tjoa & Tjoa, 2016). With funding agencies beginning to ask for an alignment of research proposals with the Sustainable Development Goals, the latter might become a common approach that could very well be picked up by conferences in the future.

Yet, my idea to ask explicitly what we strive to sustain and for whom we aim to sustain it goes further than the above approaches, in the sense that it implies the question of whether our work may incur harms to certain demographics throughout a system’s life cycle. I.e., it is easy to justify that a system potentially does some good but it requires a structural framework – similar safety arguments or threat modelling – to argue that a system does no harm.

Assessing the Solution

For our research, I would say that we strive to sustain the secure and safe operation of of smart actuation and sensing equipment that aims to optimise the distribution and use of resources. We thought specifically of use cases where very little computational resources are required to perform the control task, where distributed processing is useful but local control is needed as a safety mechanism to handle emergencies, and where equipment should be capable of receiving updates and reconfiguration over decades. Specifically, we thought of an underground irrigation system in a fruit orchard at a community farm I am involved with. With increasing water stress in Belgium, irrigating the orchard may be necessary in the near future. Yet, doing so with drip irrigation would be wasteful, and frequent disturbance of the top soil and its vegetation needs to be avoided as it forms a valuable ecosystem that prevents loss of soil moisture and erosion. With these requirements in mind, we extended an extremely light-weight processor – with a predicted per-unit price of around USD 1.- – with very strong security and safety features.

But whom does it serve, for whom do we sustain something? We envision our technology to be rolled out as a long-term investment in sectors that work towards climate mitigations. We therefore seek to minimise complexity and per-unit price, and open-sourced our designs and prototypes. Of course, we still proposed a new hardware design, which needs to be produced by suppliers that may perpetuate global power imbalances. Yet, with the open-designs and tooling, we hope that our approach lowers the entrance barrier for experimenting with the technology or to even build complex scenarios based on outdated field-programmable hardware (FPGAs) typically used in prototyping. The concern that remains is that our technology may not become easily available to rural communities in the global south that are most vulnerable and most affected to climate change. There is also the concern that, depending on the application domain, distributed control in this context unnecessarily exposes infrastructure to network-level attacks and failures. Our work does, however, put an emphasis on availability in local control systems, enabling the development of secure control systems that can be isolated from network-level attackers and that strive to minimise assets that need protection without preclude larger deployment scenarios such as sensor networks.

Assessing the Use Case

The potential use case from (Alder, Van Bulck, Piessens, & Mühlberg, 2021), smart farming, has been investigated regarding sustainability properties in related work. In a critical reflection of smart farming, Streed et al. (Streed, Kantar, Tomlinson, & Raghavan, 2021) conclude that “agricultural system planners need to think in a way that holistically addresses all the services that society desires, not marginally improve a system that fails to deliver all of these service […]. Advances in data collection and analytic techniques provide a valuable opportunity to re-envision agriculture in ways that have never before been tried, that more closely mimic nature (Thrall, 2013), or that maximise robotics and technical solutions. There is a need to combine computing and human action and desires into agriculture, but it should be done wisely, not ‘smart’ly.” These observations are in-line with the latest IPCC WG3 report, reporting that “Taken together, [precision agriculture] technologies provide farmers with a decision-support system in real time for the whole farm. Arguably, the world could feed the projected rise in population without radical changes to current agricultural practices if food waste can be minimised or eliminated. Digital technologies will contribute to minimising these losses through increased efficiencies in supply chains, better shipping and transit systems, and improved refrigeration.” (Grubb et al., 2022) Following these analysis, our work aims to provide secure and safe ICT support for socio-technical systems that involve high agroecological complexity (Raghavan et al., 2016) such as permaculture.

Abuse Cases and Harmful Impacts

Beyond our best intentions, we cannot avoid the following harmful abuse cases of our work: (1) implementing systems that facilitate oppression such as sensing and actuation in surveillance, policing, or migration control; (2) hiding unintended code, e.g., malware, in secured applications; (3) abuse of security features as digital rights management or to impede update, repair, or reuse of systems. These challenges can only be overcome by regulation and approaches to community governance of digital infrastructures.

Conclusions

For years already we find ourselves in a permacrisis of wars, humanitarian emergencies, and climate and ecological collapse. The IPCC’s sixth assessment report “Climate Change 2022: Impacts, Adaptation and Vulnerability” issues a stern warning: “The cumulative scientific evidence is unequivocal: Climate change is a threat to human well-being and planetary health. Any further delay in concerted anticipatory global action on adaptation and mitigation will miss a brief and rapidly closing window of opportunity to secure a liveable and sustainable future for all.” (Pörtner et al., 2022) In this context, I feel that a strong notion of environmental sustainability should form the baseline of a system’s safety requirements. That is, if “concerted anticipatory global action on adaptation and mitigation” to the climate and ecological emergency is not an inherent goal of the system, then using the system does involve unacceptable risks to human wellbeing, and the system should no longer have a place in our society. Working towards these goals, I argue that we as researchers, engineers, and as society as a whole, need to follow a path of critical refusal, reflecting on digital infrastructures, their importance for human wellbeing.

Sustainability for Safety and Security

ICT systems that sustainably incorporate safety and security are designed to minimise the risk of hazards for all parties involved in the life cycle of the system, while maximising the safe and secure life span and possibilities for reuse of that system and its components.

With respect to ICTs, and to security and safety in particular, sustainability concerns must always be reflected in the context of the systems and societal function that we aim to implement or move into the digital world. When designing digital infrastructures we must never compromise safety and security (and neither privacy) and the equitable and just implementation of – and access to – these (cf. design justice (Costanza-Chock, 2020), in favour of narrow notions of sustainability. However, if we accept sustainability as a multidimensional “active condition of problem solving, not a passive consequence of consuming less” (Tainter, 2006), we see a large window of possible design choices, not only technological ones, opening up. This leaves us with the problem of complexity regarding system design, system costs, and the impact of safety and security decisions on the system, its users, and other stakeholders. Tainter (Tainter, 2006) hints towards a number of approaches to deal with this complexity. These range from “don’t solve the problem” over approaches that “shift and defer costs” of complexity, to ideas that “connect costs and benefits” and “revolutionise the activity”. All may be valid, and – specifically regarding safety, security, and privacy – established risk assessment methodologies allow us to adequately assess complexity and potential harms for different pathways, and to guide decision making.

As an attempt to give a definition I propose the following: ICT systems that sustainably incorporate safety and security are designed to minimise the risk of hazards for all parties involved in the life cycle of the system, while maximising the safe and secure life span and possibilities for reuse of that system and its components.

… and what to do with that definition?

There are numerous additional terms, technologies and concepts that I could possibly mention to make my definition more concrete: from design justice to open-sourcing hardware and software, to recommendations for implementation and verification strategies. None of these concepts would be directly applicable.

Volume of data/information created, captured, copied, and consumed worldwide from 2010 to 2025. Source: statista.com
Volume of data/information created, captured, copied, and consumed worldwide from 2010 to 2025. Source: statista.com

However, because maintaining (i.e., sustaining) security and safety of a system will necessarily increase the overall environmental footprint of that system, we should strive to minimise the number of assets that actually require protection:8 More than 80% of all data ever created, consumed, and stored has been accumulated in just the last five years.9 As our society becomes obsessed with data collection, i.e. the production of new digital assets, many of which will eventually be compromised,10 we are missing out on opportunities do build more frugal infrastructures that can be sustainably secured and that sustainably protect people and communities. Therefore, the key to decisions about what systems we implement and how we regulate these systems must be democratic processes that reflect on critical societal functions and their allocated resources within the planetary boundaries.

Importantly, following the above definition and ideas for sustaining safety and security does not ignore industrial practice but instead asks for revised approaches to requirements engineering, engineering practices and regulatory frameworks and industry standards that consider safety and security risks beyond immediate hazards but follow inclusive life-cycle assessments of technologies and systems.

World's Biggest Data Breaches & Hacks. Source: informationisbeautiful.net
World’s Biggest Data Breaches & Hacks. Source: informationisbeautiful.net

Then, following ideas from research in software sustainability (Venters et al., 2018; Duboc et al., 2019) and devising equipment that generally has a low materials and energy footprint, that can be repurposed, that, for example, features updated security algorithms (crypto agility (Ott, Peikert, & participants, 2019)) and cryptographic credentials to enable strong notions of post-compromise security (Cohn-Gordon, Cremers, & Garratt, 2016), and generally following the principles laid out by Paverd et al. in (Paverd et al., 2019) can help us to sustain the secure and safe operation of critical infrastructures far beyond of what is currently feasible. Security and safety will still not be ecologically sustainable in isolation but they can be tamed to serve communities to guarantee people’s wellbeing and to operate within the planetary boundaries. With ICT being a pervasive industry sector like no other, one that has dramatically shaped the way we work and interact socially, that also contributes tremendously to the perpetuation of economic growth and exploitation, we must reject notions of technological development that do not subject their methods and outcomes to a reflection in the context of international climate- and intersectional justice, and that do not promote equitable access to technological advancement.

Future Directions

This paper has mostly asked question to guide research that seeks to reflect and embed sustained safety and security in the life cycle of sustainable ICT products. Next steps in this direction should be based on case studies – products or research – and seek to develop concrete criteria to assess if and how just and sustained notions of safety and security are implemented.

Acknowledgements

This paper is a follow-up of the author’s keynote on “Sustainable Security: What do we sustain, and for whom?” at the Workshop on Sustainability in Security, Security for Sustainability at DATE 2022.11 I would like to thank Birgit Penzenstadler, Christoph Becker, and Jay Chen for their constructive feedback on this paper. I also thank everyone who contributed to the discussion and development of the topic at DATE 2022 and also at the SICT Summer Schools on Sustainable ICT in 2020 and 2021.12 This research is partially funded by the Research Fund KU Leuven, the Flemish Research Programme Cybersecurity.

References

  1. Grubb, M., Chukwumer, O., Dhakal, S., Minx, J. C., L., T. F., Keywan, R., Schaeffer, R., et al. (2022). Climate change 2022: Mitigation of Climate Change.
  2. Gleirscher, M., Johnson, N., Karachristou, P., Calinescu, R., Law, J., & Clark, J. (2022). Challenges in the safety-security co-assurance of collaborative industrial robots. In The 21st Century Industrial Robot: When Tools Become Collaborators (pp. 191–214). Springer.
  3. Pörtner, H. O., Roberts, D. C., Adams, H., Adler, C., Aldunce, P., Ali, E., Begum, R. A., et al. (2022). “Climate change 2022: impacts, adaptation and vulnerability.” Retrieved from https://www.ipcc.ch/report/ar6/wg2/
  4. Cucchietti, F., Moll, J., Esteban, M., Reyes, P., & García Calatrava, C. (2022). “Carbolytics, an analysis of the carbon costs of online tracking.” Retrieved from http://carbolytics.org/report.html
  5. Wu, Y., Edwards, W. K., & Das, S. (2022). “SoK: Social Cybersecurity.” IEEE Symposium on Security and Privacy (Oakland). Retrieved from https://sauvikdas.com/uploads/paper/pdf/36/file.pdf
  6. Scopelliti, G., Pouyanrad, S., Noorman, J., Alder, F., Piessens, F., & Mühlberg, J. T. (2021). “POSTER: An Open-Source Framework for Developing Heterogeneous Distributed Enclave Applications.” In CCS ’21. New York, NY, USA: ACM. doi:10.1145/3460120.3485341
  7. Alder, F., Van Bulck, J., Piessens, F., & Mühlberg, J. T. (2021). “Aion: Enabling Open Systems through Strong Availability Guarantees for Enclaves.” In CCS ’21. New York, NY, USA: ACM. doi:10.1145/3460120.3484782
  8. Theilen, J. T., Baur, A., Bieker, F., Ammicht Quinn, R., Hansen, M., & González Fuster, G. (2021). “Feminist data protection: an introduction.” Internet Policy Review, 10(4). doi:10.14763/2021.4.1609
  9. Herzog, B., Reif, S., Preis, J., Schröder-Preikschat, W., & Hönig, T. (2021). “The Price of Meltdown and Spectre: Energy Overhead of Mitigations at Operating System Level.” In Proceedings of the 14th European Workshop on Systems Security (pp. 8–14). Online United Kingdom: ACM. doi:10.1145/3447852.3458721
  10. Streed, A., Kantar, M., Tomlinson, B., & Raghavan, B. (2021). “How Sustainable is the Smart Farm?” Workshop on Computing within Limits. doi:10.21428/bf6fb269.f2d0adaf
  11. Frick, V., Matthies, E., Thøgersen, J., & Santarius, T. (2021). “Do online environments promote sufficiency or overconsumption? Online advertisement and social media effects on clothing, digital devices, and air travel consumption.” Journal of Consumer Behaviour, 20(2), 288–308. doi:10.1002/cb.1855
  12. Millward-Hopkins, J., Steinberger, J. K., Rao, N. D., & Oswald, Y. (2020). “Providing decent living with minimum energy: A global scenario.” Global Environmental Change, 65, 102168. doi:10.1016/j.gloenvcha.2020.102168
  13. Costanza-Chock, S. (2020). Design justice: community-led practices to build the worlds we need. Information policy. Cambridge, MA: The MIT Press.
  14. Cifor, M., Garcia, P., Cowan, T. L., Rault, J., Sutherland, T., Chan, A., Rode, J., et al. (2019). “Feminist data manifest-no,” 119. Retrieved from https://www.manifestno.com/
  15. Paverd, A., Völp, M., Brasser, F., Schunter, M., Asokan, N., Sadeghi, A.-R., Esteves-Veríssimo, P., et al. (2019). “Sustainable Security & Safety: Challenges and Opportunities.” 4th International Workshop on Security and Dependability of Critical Embedded Real-Time Systems (CERTS 2019). Retrieved from http://www.icri-cars.org/wp-content/uploads/2019/01/s3-vision.pdf
  16. Alhubaiti, O., & El-Alfy, E.-S. M. (2019). “Impact of Spectre/Meltdown Kernel Patches on Crypto-Algorithms on Windows Platforms.” In 2019 International Conference on Innovation and Intelligence for Informatics, Computing, and Technologies (3ICT) (pp. 1–6). Sakhier, Bahrain: IEEE. doi:10.1109/3ICT.2019.8910282
  17. Duboc, L., Betz, S., Penzenstadler, B., Akinli Kocak, S., Chitchyan, R., Leifler, O., Porras, J., et al. (2019). “Do we Really Know What we are Building? Raising Awareness of Potential Sustainability Effects of Software Systems in Requirements Engineering.” In 2019 IEEE 27th International Requirements Engineering Conference (RE) (pp. 6–16). Jeju Island, Korea (South): IEEE. doi:10.1109/RE.2019.00013
  18. Ott, D., Peikert, C., & participants, other workshop. (2019). “Identifying Research Challenges in Post Quantum Cryptography Migration and Cryptographic Agility.” arXiv:1909.07353 [cs]. Retrieved from http://arxiv.org/abs/1909.07353
  19. Clemm, C., Nissen, N. F., Schischke, K., Dimitrova, G., Marwede, M., & Lang, K.-D. (2019). “Implications of the Circular Economy for Electronic Products.” In A. H. Hu, M. Matsumoto, T. C. Kuo, & S. Smith (Eds.), Technologies and Eco-innovation towards Sustainability I (pp. 91–104). Singapore: Springer Singapore. doi:10.1007/978-981-13-1181-9_8
  20. Lipp, M., Schwarz, M., Gruss, D., Prescher, T., Haas, W., Mangard, S., Kocher, P., et al. (2018). Meltdown. ArXiv e-prints.
  21. Kocher, P., Genkin, D., Gruss, D., Haas, W., Hamburg, M., Lipp, M., Mangard, S., et al. (2018). Spectre Attacks: Exploiting Speculative Execution. ArXiv e-prints.
  22. Mühlberg, J. T., & Van Bulck, J. (2018). “Reflections on post-Meltdown trusted computing: A case for open security processors.” ;login: the USENIX magazine, 43(3), 1–4. Retrieved from https://www.beetzsee.de/posts/papers/2018-login-meltdown.pdf
  23. Venters, C. C., Capilla, R., Betz, S., Penzenstadler, B., Crick, T., Crouch, S., Nakagawa, E. Y., et al. (2018). “Software sustainability: Research and practice from a software architecture viewpoint.” Journal of Systems and Software, 138, 174–188. doi:10.1016/j.jss.2017.12.026
  24. Prout, A., Arcand, W., Bestor, D., Bergeron, B., Byun, C., Gadepally, V., Houle, M., et al. (2018). “Measuring the Impact of Spectre and Meltdown.” In 2018 IEEE High Performance extreme Computing Conference (HPEC) (pp. 1–5). Waltham, MA: IEEE. doi:10.1109/HPEC.2018.8547554
  25. Maene, P., Götzfried, J., de Clercq, R., Müller, T., Freiling, F., & Verbauwhede, I. (2017). Hardware-Based Trusted Computing Architectures for Isolation and Attestation. IEEE Transactions on Computers, PP(99), 1–1.
  26. Eder-Neuhauser, P., Zseby, T., Fabini, J., & Vormayr, G. (2017). “Cyber attack models for smart grid environments.” Sustainable Energy, Grids and Networks, 12, 10–29. doi:10.1016/j.segan.2017.08.002
  27. Raworth, K. (2017). Doughnut economics: seven ways to think like a 21st century economist. White River Junction, Vermont: Chelsea Green Publishing.
  28. Cao, H., Gangakhedkar, S., Ali, A. R., Gharba, M., & Eichinger, J. (2016). A 5G V2X testbed for cooperative automated driving. In 2016 IEEE Vehicular Networking Conference (VNC) (pp. 1–4). IEEE.
  29. Cohn-Gordon, K., Cremers, C., & Garratt, L. (2016). “On Post-compromise Security.” In 2016 IEEE 29th Computer Security Foundations Symposium (CSF) (pp. 164–178). Lisbon: IEEE. doi:10.1109/CSF.2016.19
  30. Raghavan, B., Nardi, B., Lovell, S. T., Norton, J., Tomlinson, B., & Patterson, D. J. (2016). “Computational Agroecology: Sustainable Food Ecosystem Design.” In Proceedings of the 2016 CHI Conference Extended Abstracts on Human Factors in Computing Systems (pp. 423–435). San Jose California USA: ACM. doi:10.1145/2851581.2892577
  31. Tjoa, A. M., & Tjoa, S. (2016). “The Role of ICT to Achieve the UN Sustainable Development Goals (SDG).” In F. J. Mata & A. Pont (Eds.), ICT for Promoting Human Development and Protecting the Environment (Vol. 481, pp. 3–13). Cham: Springer International Publishing. doi:10.1007/978-3-319-44447-5_1
  32. Becker, C., Betz, S., Chitchyan, R., Duboc, L., Easterbrook, S. M., Penzenstadler, B., Seyff, N., et al. (2016). “Requirements: The Key to Sustainability.” IEEE Software, 33(1), 56–65. doi:10.1109/MS.2015.158
  33. Scott, W. (2015). Public Understanding of Sustainable Development: Some Implications for Education. International Journal of Environmental and Science Education, 10(2), 235–246.
  34. Shostack, A. (2014). Threat modeling: Designing for security. John Wiley & Sons.
  35. Penzenstadler, B., Raturi, A., Richardson, D., & Tomlinson, B. (2014). “Safety, Security, Now Sustainability: The Nonfunctional Requirement for the 21st Century.” IEEE Software, 31(3), 40–47. doi:10.1109/MS.2014.22
  36. Thrall, P. H. (2013). “Darwinian Agriculture: how understanding evolution can improve agriculture R.Ford Denison 2012. Princeton University Press.” Evolutionary Applications, 6(2), 408–410. doi:10.1111/eva.12029
  37. Gürses, S., Troncoso, C., & Diaz, C. (2011). Engineering privacy by design. In Computers, Privacy & Data Protection (CPDP).
  38. Krumdieck, S. (2011). “The Survival Spectrum: The Key to Transition Engineering of Complex Systems.” In Volume 4: Energy Systems Analysis, Thermodynamics and Sustainability; Combustion Science and Engineering; Nanoengineering for Energy, Parts A and B (pp. 401–411). Denver, Colorado, USA: ASMEDC. doi:10.1115/IMECE2011-65891
  39. Rockström, J., Steffen, W., Noone, K., Persson, Å., Chapin, F. S., Lambin, E., Lenton, T. M., et al. (2009). “Planetary Boundaries: Exploring the Safe Operating Space for Humanity.” Ecology and Society, 14(2). Retrieved from https://www.jstor.org/stable/26268316
  40. Council, N. R., & others. (2007). Software for dependable systems: Sufficient evidence? National Academies Press.
  41. Simpson, A. (2007). On ethnographic refusal: Indigeneity, ’voice’ and colonial citizenship. Junctures: The Journal for Thematic Dialogue, (9).
  42. Hilty, L. M., Arnfalk, P., Erdmann, L., Goodman, J., Lehmann, M., & Wäger, P. A. (2006). “The relevance of information and communication technologies for environmental sustainability – A prospective simulation study.” Environmental Modelling & Software, 21(11), 1618–1629. doi:10.1016/j.envsoft.2006.05.007
  43. Tainter, J. A. (2006). “Social complexity and sustainability.” Ecological Complexity, 3(2), 91–103. doi:10.1016/j.ecocom.2005.07.004
  44. Köhler, A., & Erdmann, L. (2004). “Expected Environmental Impacts of Pervasive Computing.” Human and Ecological Risk Assessment: An International Journal, 10(5), 831–852. doi:10.1080/10807030490513856
  45. Griese, H., Schischke, K., Reichl, H., & Stobbe, L. (2004). “Sustainable development of microelectronic technology processes integration of ecodesign.” In Proceedings of the Sixth IEEE CPMT Conference on High Density Microsystem Design and Packaging and Component Failure Analysis (HDP ’04) (pp. 154–159). Shanghai, China: IEEE. doi:10.1109/HPD.2004.1346690
  46. Daemen, J., & Rijmen, V. (1999). AES proposal: Rijndael.
  47. Goldschlag, D., Reed, M., & Syverson, P. (1999). Onion routing. Communications of the ACM, 42(2), 39–41.
  48. Brundtland, G. H. (1987). “Our Common Future—Call for Action.” Environmental Conservation, 14(4), 291–294. doi:10.1017/S0376892900016805

Last modified: 2023-11-10 18:01:25 +0100

worldwide from 2010 to 2025”, 2022-05-05. https://www.statista.com/statistics/871513/worldwide-data-created/

  1. “Sunday 3 April no more NS trains”, 2022-04-03 19:07 CEST. Archived at https://web.archive.org/web/20220404172602/https://www.ns.nl/en/travel-information/calamities/sunday-3-april-no-more-ns-trains.html ↩

  2. “Inside the Cunning, Unprecedented Hack of Ukraine’s Power Grid”, 2016-03-03. https://www.wired.com/2016/03/inside-cunning-unprecedented-hack-ukraines-power-grid/ ↩

  3. “Why industry is going green on the quiet”, 2019-09-08. https://www.theguardian.com/science/2019/sep/08/producers-keep-sustainable-practices-secret ↩

  4. Tainter (Tainter, 2003) was likely the first to ask these questions to reason about sustainability efforts: “Directing sustainability efforts in productive directions, then, requires understanding that it is a matter of values, not invariant biophysical processes. Some people and some ecosystems benefit from sustainability efforts, while others don’t. When confronted with the term ‘sustainability,’ therefore, one should always ask: Sustain what, for whom, for how long, and at what cost?” Benessia and Funtowicz (Benessia & Funtowicz, 2015) use this framework to analyse “ways in which techno-science modifies and determines the object and the subject of sustainability.” ↩

  5. The IPCC WG3 AR6 report suggests that net-zero industries are challenging but feasible is feasible, specifically regarding greenhouse gas emissions, requiring coordinated action: “Net-zero CO2 emissions from the industrial sector are challenging but possible. Reducing industry emissions will entail coordinated action throughout value chains to promote all mitigation options, including demand management, energy and materials efficiency, circular material flows, as well as abatement technologies and transformational changes in production processes.” (Grubb et al., 2022) ↩

  6. “Life cycle management of laptops and desktops”, 2022-04-03. https://admin.kuleuven.be/icts/english/sustainability/laptops-desktops ↩

  7. Meanwhile we have an extended article that describes this infrastructure ad the use case: (Scopelliti et al., 2022) ↩

  8. The latest IPCC WG3 report acknowledges this in a more general context: “Demand-side mitigation encompasses changes in infrastructure use, end-use technology adoption, and socio-cultural and behavioural change. Demand-side measures and new ways of end-use service provision can reduce global GHG emissions in end use sectors by 40-70% by 2050 compared to baseline scenarios, while some regions and socioeconomic groups require additional energy and resources. Demand side mitigation response options are consistent with improving basic wellbeing for all.” (Grubb et al., 2022) ↩

  9. “Volume of data/information created, captured, copied, and consumed ↩

  10. Data breaches strongly correlate with with the growing number of assets: https://www.informationisbeautiful.net/visualizations/worlds-biggest-data-breaches-hacks/ ↩

  11. Workshop on Sustainability in Security, Security for Sustainability at DATE 2022: http://sussec22.alari.ch/ ↩

  12. SICT Summer Schools on Sustainable ICT: https://www.sictdoctoralschool.com/ ↩