In this column, we reflect on the environmental impact and broader sustainability implications of resource-demanding digital applications and services such as video streaming, VR/AR/XR and videoconferencing. We put emphasis not only on the experiences and use cases they enable but also on the “cost” of always striving for high Quality of Experience (QoE) and better user experiences. Starting by sketching the broader context, our aim is to raise awareness about the role that QoE research can play in the context of various of the United Nations’ Sustainable Development Goals (SDGs), either directly (e.g., SDG 13 “climate action”) or more indirectly (e.g., SDG 3 “good health and well-being” and SDG 12 “responsible consumption and production”).

The ambivalent role of digital technology

One of the latest reports from the Intergovernmental Panel on Climate Change (IPCC) confirmed the urgency in drastically reducing emissions of carbon dioxide and other human-induced greenhouse gas (GHG) emissions in the years to come (IPCC, 2021). This report, directly relevant in the context of SDG 13 “climate action”, confirmed the undeniable and negative human influence on global warming and the need for collective action. While the potential of digital technology (and ICT more broadly) for sustainable development has been on the agenda for some time, the context of the COVID-19 pandemic has made it possible to better understand a set of related opportunities and challenges.

 

First of all, it has been observed that long-lasting lockdowns and restrictions due to the COVID-19 pandemic and its aftermath have triggered a drastic increase in internet traffic (see e.g., Feldmann, 2020). This holds particularly for the use of videoconferencing and video streaming services for various purposes (e.g., work meetings, conferences, remote education, social gatherings, just to name a few). At the same time, the associated drastic reduction of global air traffic and other types of traffic (e.g., road traffic) with their known environmental footprint, has had undeniable positive effects for the environment (e.g., reduced air pollution, better water quality see e.g., Khan et al., 2020). Despite this potential, the environmental gains enabled by digital technology and recent advances in energy efficiency are threatened by digital rebound effects due to increased energy consumption and energy demands related to ICT (Coroamua et al., 2019Lange et al., 2020).

 

In the context of ever-increasing consumption, there has for instance been a growing focus in the literature on the negative environmental impact of unsustainable use and viewing practices such as binge-watching, multi-watching and media-multitasking, which have become more common over the last years (see e.g., Widdicks, 2019).

 

While it is important to recognize that the overall emission factor will vary depending on the mix of energy generation technologies used and region in the world (Preist et al., 2014), the above observation also fits with other recent reports and articles, which expect the energy demands linked to digital infrastructure, digital services and their use to further expand and which expect the greenhouse gass emissions of ICT relative to the overall worldwide footprint to significantly increase (see e.g., Belkhir et al., 2018Morley et al., 2018Obringer et al., 2021).

 

Hence, these and other recent forecasts show a growing and even unsustainable high carbon footprint of ICT in the middle-term future, due to, among others, the increasing energy demand of data centers (including e.g., also the energy needed for cooling) and the associated traffic (Preist et al., 2016).

 

Another set of challenges that became more apparent can be linked to the human mental resources and health involved as well as environmental effects. Here, there is a link to the abovementioned Sustainable development goals 3 (good health and well-being) and 12 (sustainable consumption and production).

 

For instance, the transition to “more sustainable” digital meetings, online conferences, and online education has also pointed to a range of challenges from a user point of view.  “Zoom fatigue” being  a prominent example illustrates the need to strike the right balance between the more sustainable character of experiences provided by and enabled through technology and how these are actually experienced and perceived from a user point of view (Döring et al., 2022Raake et al., 2022).

 

Another example is binge-watching behavior, which can in certain cases have a positive effect on an individual’s well-being, but which has also been shown to have a negative effect through e.g., feelings of guilt and goal conflicts  (Granow et al., 2018) or through problematic involvement resulting in e.g., chronic sleep issues  (Flayelle, 2020).

 

From the “production” perspective, recent work has looked at the growing environmental impact of commonly used cloud-based services such as video streaming (see e.g., Chen et al., 2020Suski et al., 2020The Shift Project, 2021) and the underlying infrastructure consisting of data centers, transport network and end devices (Preist et al., 2016Suski, 2020Preist et al., 2014). As a result, the combination of technological advancements and user-centered approaches aiming to always improve the experience may have undesired environmental consequences. This includes stimulating increased user expectations (e.g., higher video quality, increased connectivity and availability, almost zero-latency, …) and by triggering increased use, unsustainable use practices, resulting in potential rebound effects due to increased data traffic and electricity demand.

 

These observations have started to culminate into a plea for a shift towards a more sustainable and humanity-centered paradigm, which considers to a much larger extent how digital consumption and increased data demand impact individuals, society and our planet (Widdicks et al., 2019Priest et al., 2016Hazas & Nathan, 2018). Here, it is obvious that experience, consumption behavior and energy consumption are tightly intertwined.

How does QoE research fit into this picture?

This leads to the question where research on Quality of Experience and its underlying goals fit into this broader picture, to which extent related topics have gained attention so far and how future research can potentially have an even larger impact.

As the COVID-19 related examples above already indicated, QoE research, through its focus on improving the experience for users in e.g., various videoconferencing-based scenarios or immersive technology-related use cases, already plays and will continue to play a key role in enabling more sustainable practices in various domains (e.g., remote education, online conferences, digital meetings, and thus reducing unnecessary travels, …) and linking up to various SDGs.

 

A key challenge here is to enable experiences that become so natural and attractive that they may even become preferred in the future. While this is a huge and important topic, we refrain from discussing it further in this contribution, as it already is a key focus within the QoE community. Instead, in the following, we first of all reflect on the extent to which environmental implications of multimedia services have explicitly been on the agenda of the QoE community in the past, what the focus is in more recent work, and what is currently not yet sufficiently addressed. Secondly, we consider a broader set of areas and concrete topics in which QoE research can be related to environmental and broader sustainability-related concerns.

 

Traditionally, QoE research has predominantly focused on gathering insights that can guide the optimization of technical parameters and allocation of resources at different layers, while still ensuring a high QoE from a user point of view. A main underlying driver in this respect has traditionally been the related business perspective: optimizing QoE as a way to increase profitability and users/customers’ willingness to pay for better quality  (Wechsung, 2014). While better video compression techniques or adaptive video streaming may allow the saving of resources, which overall may lead to environmental gains, the latter has traditionally not been a main or explicit motivation.

 

There are however some exceptions in earlier work, where the focus was more explicitly on the link between energy consumption-related aspects, energy efficiency and QoE. The study of (Ickin, 2012) for instance, aimed to investigate QoE influence factors of mobile applications and revealed the key role of the battery in successful QoE provisioning. In this work, it was observed that energy modeling and saving efforts are typically geared towards immediate benefits of end users, while less attention was paid to the digital infrastructure (Popescu, 2018).

 

Efforts were further also made in the past to describe, analyze and model the trade-off between QoE and energy consumption (QoE perceived per user per Joule, QoEJ) (Popescu, 2018) or power consumption (QoE perceived per user per Watt, QoEW) (Zhang et al., 2013), as well as to optimize resource consumption so as to avoid sources of annoyance (see e.g., (Fiedler et al., 2016). While these early efforts did not yet result in a generic end-to-end QoE-energy-model that can be used as a basis for optimizations, they provide a useful basis to build upon.

 

A more recent example (Hossfeld et al., 2022) in the context of video streaming services looked into possible trade-offs between varying levels of QoE and the resulting energy consumption, which is further mapped to CO₂ emissions (taking the EU emission parameter as input, as this – as mentioned – depends on the overall energy mix of green and non-renewable energy sources). Their visualization model further considers parameters such as the type of device and type of network and while it is a simplification of the multitude of possible scenarios and factors, it illustrates that it is possible to identify areas where energy consumption can be reduced, while ensuring an acceptable QoE.

 

Another recent work (Herglotz et al., 2022) jointly analyzed end-user power efficiency and QoE related to video streaming, based on actual real-world data (i.e., YouTube streaming events). More specifically, power consumption was modeled and user-perceived QoE was estimated in order to model where optimization is possible. They found that optimization is possible and pointed to the importance of the choice of video codec, video resolution,frame rate and bitrate in this respect.

 

These examples point to the potential to optimize at the “production” side, however, the focus has more recently also been extended to the actual use, user expectations and “consumption” side (Jiang et al., 2021Lange et al., 2020Suski et al., 2020Elgaaied-Gambier et al., 2020) Various topics are explored in this respect, e.g., digital carbon footprint calculation at the individual level (Schien et al., 2013Preist et al., 2014), consumer awareness and pro-environmental digital habits (Elgaaied-Gambier et al., 2020Gnanasekaran et al., 2021), or impact of user behavior (Suski et al., 2020).

 

While we cannot discuss all of these in detail here, they all are based on the observation that there is a growing need to involve consumers and users in the collective challenge of reducing the impact of digital applications and services on the environment (Elgaaied-Gambier et al., 2020Priest et al., 2016).

 

QoE research can play an important role here, extending the understanding of carbon footprint vs. QoE trade-offs to making users more aware of the actual “cost” of high QoE. A recent interview study with digital natives conducted by some of the co-authors of this column  (Gnanasekaran et al., 2021) illustrated that many users are not aware of the environmental impact of their use behavior and expectations, and that even with such insights, substantial drastic changes in behavior cannot be expected.

 

The lack of technological understanding, public information and social awareness about the topic were identified as important factors. It is therefore of utmost importance to trigger more awareness and help users see and understand their carbon footprint related to e.g., the use of video streaming services (Gnanasekaran et al., 2021). This perspective is currently missing in the field of QoE and we argue here that QoE research could – in collaboration with other disciplines and by integrating insights from other fields – play an important role here.

 

In terms of the motivation for adopting pro-environmental digital habits, Gnanasekaran et al., (2021) found that several factors indirectly contribute to this goal, including the striving for personal well-being. Finally, the results indicate some willingness to change and make compromises (e.g., accepting a lower video quality), albeit not an unconditional one: the alignment with other goals (e.g., personal well-being) and nature of the perceived sacrifice and its impact play a key role.

 

A key challenge for future work is therefore to identify and understand concrete mechanisms that could trigger more awareness amongst users about the environmental and well-being impact of their use of digital applications and services, and those that can further motivate positive behavioral change (e.g., opting for use practices that limit one’s digital carbon footprint, mindful digital consumption). By investigating the impact of various more environmentally-friendly viewing practices on QoE (e.g., actively promoting standard definition video quality instead of HD, nudging users to switch to audio-only when a service like YouTube is used as background noise or stimulating users to switch to the least data demanding viewing configuration depending on the context and purpose), QoE research could help to bridge the gap towards actual behavioral change.

Final reflections and challenges for future research

We have argued that research on users’ Quality of Experience and overall User Experience can be highly relevant to gain insights that may further drive the adoption of new, more sustainable usage patterns and that can trigger more awareness of implications of user expectations, preferences and actual use of digital services. However, the focus on continuously improving users’ Quality Experience may also trigger unwanted rebound effects, leading to an overall higher environmental footprint due to increased use of digital applications and service. Further, it may have a negative impact on users’ long-term well-being as well.

We therefore need to join efforts with other communities to challenge the current design paradigm from a more critical stance, partly as “it’s difficult to see the ecological impact of IT when its benefits are so blindingly bright” (Borning et al., 2020). Richer and better experiences may lead to increased, unnecessary or even excessive consumption, further increasing individuals’ environmental impact and potentially impeding long-term well-being. Open questions are therefore: Which fields and disciplines should join forces to mitigate the above risks? And how can QoE research — directly or indirectly — contribute to the triggering of sustainable consumption patterns and the fostering of well-being?

 

Further, a key question is how energy efficiency can be improved for digital services such as video streaming, videoconferencing, online gaming, etc., while still ensuring an acceptable QoE. This also points to the question which compromises can be made in trading QoE against its environmental impact (from “willingness to pay” to “willingness to sacrifice”), under which circumstances and how these compromises can be meaningfully and realistically assessed. In this respect, future work should extend the current modeling efforts to link QoE and carbon footprint, go beyond exploring what users are willing to (more passively) endure, and also investigate how users can be more actively motivated to adjust and lower their expectations and even change their behavior.

These and related topics will be on the agenda of the Dagstuhl seminar  23042 “Quality of Sustainable Experience” and the conference QoMEX 2023 “Towards sustainable and inclusive multimedia experiences”.

Credits:

Authors: Katrien De Moor (NTNU), Markus Fiedler (BTH), Alexander Raake (TU Ilmenau), Ashok Jhunjhunwala (Indian Institute of Technology), Vahiny Gnanasekaran (NTNU), Sruti Subramanian (NTNU), Thomas Zinner (NTNU)
Editors: Tobias Hoßfeld (University of Würzburg, Germany), Christian Timmerer (Alpen-Adria-Universität (AAU) Klagenfurt and Bitmovin Inc., Austria)
October 10th, 2022
Read more at:
https://records.sigmm.org/2022/10/07/towards-the-design-and-evaluation-of-more-sustainable-multimedia-experiences-which-role-can-qoe-research-play/