For a long time, those involved in XR and digital creation have been sailing at a dead end: what is the environmental impact of digital equipment or infrastructures? What is the footprint of a work, from production to distribution? These are particularly complex questions, given the many variables that need to be integrated into calculation methods. And yet, these impacts need to be estimated, because the underlying problem is so powerful: is XR really compatible with environmental issues? A number of recent announcements offer a first glimpse of the picture. Here’s an overview, punctuated by testimonials from industry professionals.
Cover: FAR AWAY @ CHRONIQUES, photo Pierre-Gondard
The weight of digital technology
Writing an article on eco-responsibility and XR is a tricky and ambitious subject. Firstly, it’s difficult because there are few studies and impact reports to enrich scientific knowledge and current debates. At least this was true until the recent release, in March 2024, of the CEPIR (Cas d’Étude pour un Immersif Responsable, in French) report, whose work (carbon footprints and equipment lifecycle analyses…), discussed later in this article, provides data on the environmental impacts of XR (focused on VR/AR/MR). And then there’s the scope, as the preliminary to any analysis involves untying a few knots and tackling a few contextual issues. It will have escaped no one’s notice that, since the Paris agreements, several countries, including the EU member states, have committed themselves to achieving carbon neutrality, i.e. a balance between greenhouse gas (GHG) emissions and absorptions. This is an ambitious goal given the current situation: in 2019, the digital sector accounted for around 3.8% of global GHG emissions (In French – GreenIT). In 2022 in France, this figure was estimated at 2.5% of the country’s carbon footprint (In French – ADEME), slightly more than the waste sector estimated at 2% of GHGs.
And the trend is set to increase: ARCEP (In French) anticipates a scenario for 2040 in which the digital sector accounts for 6.7% of French GHG emissions, i.e. 4.2 points more than in 2022. These figures help us to understand the scale and weight of digital technology, without actually making it a reality. Digital equipment provides a small part of the answer: in 2019, there was 3.5 billion smartphones worldwide, 3.1 billion TVs, screens and projectors, 1.1 billion DSL boxes and 67 million hosted servers (In French – GreenIT). The number of VR/AR headsets in circulation worldwide, on the other hand, is harder to estimate. According to various sources, it’s probably in the tens of millions. All these figures are essential to provide an initial framework for the subject, and a perspective on future developments. All the more so as advances in generative AI combined with XR are set to drive an explosion in usage, and with it the GHGs emitted (increased infrastructure, GPUs, etc.). A rough estimate: a rise of around 80% in global GHG emissions over the next few years (The AI Threats to Climate Change).
GHGs, a limited metric
However, we must accept the premise that the environmental footprint is measured in terms of greenhouse gas emissions, even though they are only a relative indicator. In fact, scientists favor a transversal approach and an analysis grid based on the 9 planetary limits, of which climate is only one component (among others: erosion of biodiversity; disruption of the nitrogen and phosphorus cycles; changes in land use; the freshwater cycle, etc.). It should also be pointed out that decarbonization ambitions also lead to strong political orientations. Firstly, the origins of energy production, whether nuclear or not (in France, it accounts for a large proportion of the energy mix).
However, as Amaury La Burthe, co-founder of Novelab (which he has since left) and one of the initiators of CEPIR, points out: “Carbon is already a well-documented and reliable datum. We know the average emission factor for car travel in France, or that for electricity consumption. Although incomplete for representing the diversity of environmental impacts, it is an indicator that can be used to initiate a change of trajectory for the coming years.” So, the scene is set: yes, digital technology does represent a considerable proportion of the world’s ecological footprint (a proportion that will be even greater in the years to come). Yes, the issue of impact is systemic and goes beyond the simple question of digital technology. And yes, the data presented will undoubtedly need to be consolidated by new calculation methods, incorporating variables other than GHGs.
First numbers on the carbon footprint of an XR artwork
Until recently, very few studies were able to draw any structuring conclusions. Here are two particularly interesting studies. The first dates back to 2022, and was initiated by Stereolux (through its Labo Arts & Techs) and the Chevalvert studio, which specializes in interactive design. The aim was to use the FAR AWAY installation (composed of 12 luminous and kinetic totems reacting to audience movements) to measure the contribution of a digital work to climate change, and also to other resources (water, minerals, etc.). A particular feature of this study is that it takes into account the life cycle of the work, i.e. its manufacture (analysis of components and their provenance), its operation (electricity consumption) and its distribution at festivals and events (transport, assembly, dismantling, etc.). Some instructive data emerge from the summary (In French – to be read in full). In particular, the importance of electronic components (LEDs, circuit boards, sensors, etc.) in the environmental impact. “I was astonished by the impact that very small parts can have. Electronic components only represent around 1.5% of the total mass of the work, but contribute between 18% and 65% of the environmental impact, which is considerable. Almost all of these components are manufactured in China, which greatly increases their impact (transport, energy mix, etc.). Even components sold as European are, admittedly, assembled in Europe, but the parts still come from China. So, to my knowledge, there is no more ecologically sober supply solution at the moment“, says Stéphane Buellet, co-founder of studio Chevalvert (In French – HACNUMedia). Another important fact: the installation’s second GHG footprint comes from transporting the work and distributing it. It is estimated at around 3 tCO₂e over its life cycle, the equivalent of about 13,000 kilometers by car.
The second CEPIR study (Cas d’Étude pour un Immersif Responsable), revealed in March 2024 and driven by Tiny Planets, Novelab and La Coopérative Carbone, provides new insights into the environmental footprint caused by the design-production-distribution of an XR work. Two VR works were studied between 2019 and 2022. The method for calculating ecological impacts was established after the design-production-distribution of UMAMI, but before that of OKAWARI. The data collected from this last work are particularly interesting: excluding festival-goers’ travel (the installation was programmed at Venice Immersive, Lumières sur le Quai, IFDA and Cité des Sciences), the 4 main GHG emissions are broken down as follows: 26% for energy consumption and buildings. 26% on accommodation and catering for team members. 20% on travel, 18% on digital and IT. The remainder is divided between various operating expenses. The CEPIR study nevertheless incorporates a more original method of calculation, by adding the travel costs of festival-goers who have seen/experienced OKAWARI. Amaury La Burthe explains that “if a festival-goer came from Montreal and saw 10 works, we attributed 1/10 of their festival footprint to the impact of the work.” This new perimeter – distinguishing between carbon balance and carbon footprint (the sum of emissions produced and emissions linked to products imported and consumed) – gives a new coloring to the data. As a result, festival-goers’ travel accounts for 78% of the work’s carbon footprint. This means a total of 53 t CO₂e for OKAWARI’s complete footprint, divided between the design and manufacture of the work equivalent to 12 tCO₂e and festival distribution equivalent to 41 tCO₂e . Remember that if the annual individual footprint of French people is around 8 tCO₂e for the year 2022 (In French – ADEME), OKAWARI has the same GHG footprint as those of 6 French people over one year.
What about hardware?
Alongside this analysis, CEPIR’s deliverables include a study of the life cycle assessments (LCAs) of XR-related equipment, primarily VR headsets. CEPIR experts dismantled 4 headsets and their controllers, and measured the various components according to 16 environmental criteria (climate change, depletion of abiotic resources, acidification, freshwater ecotoxicity, human toxicity, fine particle emissions, etc.). To avoid possible disputes and legal action, these helmets have been anonymized. Nevertheless, one differentiating factor stands out: the technology used. Some VR headsets are equipped with LCD screens, others with OLED screens. A few preliminary findings on components: the LED panel accounts for the majority of a headset’s environmental impact. Additional emission factors, such as batteries, VR controllers or terminals (internet box, computer, etc.) were also investigated with multiple gCO2e/hour data. This also represents a complex calculation in terms of carbon-based and non-carbon-based electricity generation, and once again reveals major disparities between countries around the globe. In fine, here are the two facts to remember about the carbon footprint of equipment manufacturing and distribution: an LCD screen VR headset (as well as 2 controllers) represents almost 50 kgCO₂e/pack. An OLED screen VR headset (as well as 2 controllers) represents 92 kgCO₂e/pack. What’s more, each VR headset and its controllers consume on average the equivalent of 27 liters of oil and around 100 m3 of water on average, equivalent to the annual consumption of two French people. It should be noted that the figures given by Apple, concerning the latest Apple Vision pro, are still well beyond the CEPIR data estimates and figure at around 335kgCO₂e in total, including 207.7 for manufacturing alone (Apple). This is probably due to the definition of the screens and the large number of technologies integrated into this headset.
These figures are probably not very meaningful without contextualizing them and cross-referencing them with future scenarios. Based on scientific reports (ITU), CEPIR estimates that the individual carbon budget in 2030 for digital and to maintain sustainability objectives is 53 kgCO₂e/inhabitant/year all inclusive (terminals, data centers and networks)! A scenario in which the individual budget is the same worldwide. In other words, a consumer purchasing a single LCD VR headset in 2030 would see his or her digital carbon credit used up for an entire year.
More bouncing-back effects to come
The problem with this equipment is compounded by the obsolescence of materials, the difficulty of maintenance and the lack of recycling channels. “Headsets have too short a lifespan, and the technology has not yet reached a stage of maturity: it is constantly evolving, and new equipment is released every 2 to 4 years. Some XR players are trying to repair and maintain their stock of used helmets using parts salvaged from other helmets, but there’s no way of getting spare parts. It’s not going very far, because there’s no structure to it,” comments Marc Lopato, co-founder of Diversion cinema, a broadcaster and distributor of VR works. Other rebound effects can also be observed. Starting with infrastructures. Laurent Chicoineau, director of the Quai des Savoirs in Toulouse, which regularly hosts VR devices and is organizing the AI: Double-je exhibition in 2024, raises an interesting point. “One of the problems with digital is that infrastructure needs are not proportional to usage, they’re exponential. We need to ensure uninterrupted service in the event of network breakdowns or severed cables. Digital technology multiplies infrastructures to respond to the principle of resilience.” In other words: the more digital (XR devices, equipment, etc.), the more infrastructure. Finally, certain ecological footprint items continue even after an XR work has stopped being broadcast. This is a paradox pointed out by Amaury La Burthe and Landia Egal (co-founder of Tiny Planets), who explain that “the BNF (National Library of France, which archives all contemporary cultural works) is obliged to deliver executables and store works financed by the CNC, the National Centre for Cinema and the Moving Image. This storage, this conservation, is a carbon footprint that will never stop. It’s a problem for which there are few solutions.”
An assessment… and then what?
Far from condemning XR players – all the more so as it’s still not easy to assess the social impact of the development of new collective narratives – this general observation and the studies cited set the scene for the next logical step: how do we measure the positive and negative contributions of XR? How can we steer public policies (in the digital and cultural fields, etc.) on the basis of informed choices? How can we structure the XR sector and support it in its environmental transition? What concrete actions should be taken? These are just a few of the questions addressed by several CEPIR deliverables (the report ‘La XR à l’horizon 2030. Projection into possible futures” or “The best practices guide”) already provide answers, will be the subject of a second article to be published shortly on XRMust.
Leave a Reply
You must be logged in to post a comment.