Design and interdisciplinary collaboration are crucial for promoting a greener culture within data centres. Deborah Andrews, Professor of Design for Sustainability and Circularity at London South Bank University, emphasises the need for designers to consider the environmental impact of their decisions throughout the product life cycle, and encourages further efforts into awareness, education and incentives for all current or prospective stakeholders and talent.
What does your day-to-day look like as Professor of Sustainability and Circularity at London South Bank University?
It’s quite diverse. Sometimes I’m completely immersed in research with colleagues or partners outside the university; on other days I teach, or I’m involved in enterprise projects. Whether that means I’m a renaissance woman or a jack-of-all-trades depends on your stance, but I would say that 99% of the activities that I engage in are related to sustainability in some way.
Although I’m working with people in the industry and trying to encourage change, I also work with students and encourage them to think about sustainability but also improving it; eventually it becomes reflexive and intuitive, and it’s something that they will do automatically.
How do you see the role of design in promoting sustainability and circularity within the context of data centres, and what specific design principles do you believe are crucial for achieving these goals?
We’re very aware that the sector has developed rapidly, along with the general scale of development – it’s absolutely astonishing. Emphasis has been on service provision rather than thinking about the wider implications of decisions – incremental tweaks have been made rather than really understanding first principles.
We’ve ended up designing products that are suitable for a linear economy, in other words, they’re designed for life up to its service without any consideration of what happens at end-of-life. This isn’t ultimately when products no longer usable, but when they come to mid-life, when they might no longer meet a first customer’s requirements. Design doesn’t accommodate how could they be improved or upgraded, for instance.
In terms of responsibility, I think designers have a lot. There are statistics that say the decisions made in the design phase can influence the environmental impact of a product through its life up to 80%. We need to stop making superficial tweaks and go back to first principles and start designing with end-of-life and mid-life in mind so that we can encourage product life extension, recycling and reclamation of materials to develop closed loops.
What innovative approaches or design strategies do you recommend for creating more energy efficient and environmentally friendly data centres?
I would argue that we’re doing quite a lot around operational energy efficiency already. There is a lot of encouragement because of legislation, but the biggest challenge is embodied energy and all the other impacts associated with manufacturing hardware. It’s a big challenge because we need to understand what products or equipment is made from in the first place. This will help us see whether we can substitute materials that have a higher environmental impact or change the way in which products are designed to facilitate product life extension and recycling. It’s also thinking about making products more resource-efficient initially; rather than having 10 screws, if you only need five, you put five in – you don’t need to over-engineer products.
Lots of smaller changes are going to be essential to improving sustainability across the data centre, and looking not just at the products, equipment, servers, networking equipment and so forth, but also the larger data centre infrastructure in the longer term. As well as this, how the operational energy is generated is going to be critical to overall sustainability.
Collaboration is often key in addressing complex sustainability challenges. In your experience, how can interdisciplinary collaboration between designers, engineers and other stakeholders contribute to creating more sustainable and circular data centre solutions and what role does education play in fostering this?
I think interdisciplinarity and collaboration are absolutely essential to develop any truly sustainable or circular solution – it doesn’t matter what sector you’re working in. Part of the challenge we have in the data centre industry is the speed and scale of sectoral growth. But what’s tended to happen is that a lot of people have developed incredibly fine knowledge and deep levels of skills in one particular area without really appreciating the implications of what they do in context. I’ve heard colleagues refer to the data centre industry as a sort of silo culture, where sub-sectors work in isolation and don’t really know what other people do or other sub-sectors do.
A good example of positive collaboration through the CEDaCI (Circular Economy for the Data Centre Industry) project is where we brought together representatives from as many different sub-sectors as possible to work on developing our circular data centre compass through a co-creative process. There was an unexpected benefit where members of the workshops commented they had learned as much from other people as they’d learned from us – they gained insight into what other people in other sectors do.
If you think about it, any circular economy is multifaceted; there are so many different factors which contribute to the industry as a whole. You have to take a whole system’s approach to the challenge of circularity and it’s about understanding how everything relates to everything else. Whatever activity or action you take, in one part of the network or the product lifecycle, has implications for every other part of the network and the life cycle. Collaboration, whole systems thinking and interdisciplinarity is essential.
In terms of education, it would be great to run more CPD courses for people who are already employed in the industry. One of the challenges would be to get people over the doorstep. It’s thinking about how you get people who are slightly resistant to change to come along and find out about what other people do, for example. From a university perspective, I think we’re encouraging students who are in design and engineering courses to work with students on other courses – for instance, marketing – for mutual insights into other professions. If students have had experience of interdisciplinary projects in education, hopefully they’ll carry that into the workplace and be more confident about encouraging cross-sector work in the real world.
How should eco-fatigue and eco-anxiety among data centre operators be managed and how will this help fuel a better understanding of Sustainable Development Goals (SDGs)?
This is another big challenge, and it cuts across all industries and education. It’s quite sad that we are bombarded with so much negative information that people are almost feeling as though they have to shut off. One of the things that we have tried to do with students who feel disempowered is to encourage them to recognise that the cumulative effects of many people doing one small thing can be very positive.
We’ve also tried using reverse psychology to make learning about sustainability and unsustainability really good fun. In education, you’re always encouraged to do the ‘right thing’, so we developed a project where students had to develop the most unsustainable solution to a challenge that they could possibly think of. They also got points for trashing as many sustainable development goals as possible. Initially, they couldn’t quite grasp the idea that they were being asked to do something bad, but once they got over that hurdle, they went crazy – they loved it.
We applied the principles of reverse psychology at an international life cycle management conference last year. Through a workshop, we tried to encourage white, middle-class, Western European delegates to think about the Global South and SDGs in a more sympathetic, empathetic and understanding way. Again, we got the delegates to think about the worst, most unethical things that they could possibly do if they were introducing a project in the Global South – and it worked very well. We did mention that in all circumstances these practices shouldn’t be taken up, but it was a good way of getting people to think about challenges from a different perspective. Delegates from the Global South relayed back that they enjoyed it and approved; they added real value to the comments that were made.
One of the main things we can do is to lighten the subject in a way to make people feel more empowered. The idea that every little action – when you aggregate and accumulate – can bring about real change. Whether members of the data centre industry tend to be a bit more serious than some of our students, whether they would be open to our reverse psychology approach, we have yet to see.
Circular economy principles emphasise the importance of reducing, reusing and recycling materials. How can these principles be effectively applied to the design, construction and operations of data centres?
There are similar and different challenges here. Firstly, the actual data centre building should last 60 years; much of the infrastructure has around a 20–30-year lifespan; and then we have networking equipment with 10 years. Although this is on average, as in some instances they last for a lot less. This isn’t necessarily because they don’t work but because of changes in compute technology; for example, when they’re upgraded or replaced to accommodate the latest technology.
Hyperscalers hold some examples of good practice by selling on a product to other businesses for upgrading or resale for product life extension. What we need to do is to get rid of barriers to the resale of products, because a product that comes out of a hyperscale data centre will easily meet some of the requirements of another customer who require lower specifications. There is pressure from OEMs who, understandably, want to sell more brand-new equipment, encouraging customers that their new products are more energy efficient and have more compute power. However, we should be educating customers about understanding their requirements and highlighting that most second-life products will likely meet them.
Overall, there are barriers like legislation around warranties and also secondary market products not being as good or as effective as a new product. We’ve got to stop thinking like that. We haven’t got enough resources to continue to make new equipment all the time; we haven’t got the infrastructure to recycle equipment just coming out of service or as it’s reaching its end-of-life. Ideally, we need to keep things in service for as long as technically and economically viable.
There are also issues around data security, shredding, hard drives and solid states as these tend to interfere with the recycling process – on the other hand, if something works well, why would you want to chop it up and recycle it?
How do you envision the role of adaptive and future-proof design, and ensuring that data centres remain sustainable and contribute positively to circular economy goals over the long term?
We need to encourage the industry to think longer term, and not just from product model to product model. If a building is going to last for 60 years, then we need to ensure that there’s adequate infrastructure to make sure it really does last for that period and to ensure that any kind of change can be introduced later. If your building is reliant on a particular kind of energy generation, it’s key to assess the environment and note if there’s enough land to either build a wind or solar farm, or for other alternatives and renewables you may want to incorporate in the future. It’s about taking a much more systematic approach.
Referring to resource efficiency, we need to ensure that products are designed with circular principles or not using more materials than needed. For example, some networking equipment includes a lot of plastic components – which is great when you’re thinking about reducing weight of product, as it’s better than metal – but are coated with a fireproof substance making recycling impossible. By using that one material, you’re limiting it to one life in service. It’s worth thinking about whether it’s possible to substitute some of these plastics with metals, especially as there’s already a well-established infrastructure for recycling iron, steel, copper, aluminium and gold.
Additionally, we need to think about firmware. The legal minimum, and life of, is around eight years – it could be made open source after that period for older coding to be kept in service. Circularity is thinking about all the different aspects that contribute to service: hardware, software and energy efficiency.
What are your thoughts on the talent and skills shortage the data centre industry is facing, along with the challenges of AI potentially automating some roles, and how does CEDaCI (Circular Economy for the Data Centre Industry) encourage more jobs in the industry?
One of the issues here is that an awful lot of people have absolutely no idea what happens to data. They don’t realise that when they’re sending messages to friends or looking at photographs that data is involved. Schools and colleges have limited information about what the data centre industry is and what careers are available, as well as what’s involved to get into those careers. There needs to be a lot more transparency about what the industry is and the options available.
A key factor of the industry is the virtualisation side, where AI is going to have a significant role. There are many positive things that could come out of using AI, such as identifying sustainable solutions, but at the end of the day, unless everything is roboticised, we’re still going to need technicians to go into data centres and monitor or deal with technical problems. It’s likely that some jobs will change a little from current jobs, but we’re still going to need humans to be involved in the sector and deliver human expertise.
Recently, I noticed a big operator produced a kid’s book about data centres – education from an early stage will help demystify, fuel awareness and make these careers sound exciting. Another element is getting kids involved in electronics or hands-on projects at school. While this may require investment from the government and ensuring schools are adequately equipped with the right resources, the main aim is to encourage kids to make things with their hands. It’s fine sitting in front of a keyboard and doing things virtually, but they need to be educated to use their hands and feel confident doing so, making physical artefacts in line with building those virtual products.
Education, in the longer term, has an important role to play. There’s quite a gap between primary school kids now and those who are currently working in the data centre industry. Somehow, we need to make it attractive, exciting and interesting.
You’ve encouraged further attention and investigation from researchers in the data centre industry, but what can data centre operators and managers implement themselves to make progress towards the SDGs?
This is anecdotal, but from experience like trade and industry events, there are people who get it and people who don’t – those who don’t, are reluctant to find out and they’re hard to reach. However, I would say the people who do get it are leading and beginning to make change, which is encouraging those who don’t.
The SDGs can seem really overwhelming. There are 17 in total, each with their own criteria. One of our research students recently completed her master’s based on SDGs and the data centre industry, uncovering what the industry understands about them and what they’re doing. Sadly, we found some companies are doing positive things, but what they tend to do is look at the list and tick the goals that they are already working towards or have achieved – they don’t consider the goals that aren’t directly involved with what they’re doing.
To tackle this, I think legislation, or a rewarding body would encourage the change needed. Most rewards are simply having a statement on your website, but it works – having a system that could stimulate transformation would be good for organisations to work towards and get something in return.
