Net zero now forms the heart of policies and strategies across government, from “levelling up” to “build back better”. Combined with the UK’s National Infrastructure Strategy, it is clear that engineering – particularly new technologies such as robotics and AI – has a core role in our decarbonisation efforts. But how green are the technologies that will drive the green recovery? In this blog, Professor Michael Fisher outlines how the digital revolution in engineering will affect every aspect of net zero, and the role of policymakers and regulators in ensuring these technologies are themselves striving for decarbonisation.
- Digital technologies represent nearly a third of the carbon savings the UK must make in the next decade.
- AI, robotics, and other digital technologies are being applied to everything from agriculture to energy distribution, in a drive for efficient resource management.
- However, these technologies have significant environmental costs in themselves, from production of components, to the huge energy demands of cloud computing.
- New standards are needed to ensure new technologies are developed and deployed sustainably, within existing carbon budgets.
The breadth of issues around sustainability and “net zero” demands action from all disciplines, across all sectors. Environmental sustainability issues sit within the broader framework of the United Nations Sustainable Development Goals, particularly the aim to develop sustainable production and consumption patterns and, unsurprisingly, engineering will likely play a pivotal role. It is clear that advanced technology of various forms – and more specifically new technologies such as AI, robotics, and broader digital technologies – might be able to provide vital parts of our drive towards energy efficiency and a reduction in waste.
Artificial Intelligence (AI)
The UK Prime Minister’s “Ten Point Plan for a Green Industrial Revolution” highlights the development of “disruptive technologies such as artificial intelligence for energy” as a key priority. Furthermore, the UK’s National AI Strategy suggests a range of ways in which AI could help, using: “machine learning to forecast electricity generation and demand and control its distribution around the network; data analysis to find inefficiencies in emission-heavy industries; and AI to model complex systems, like Earth’s own climate”. More generally, “AI is increasingly seen as a critical technology to scale and enable these significant emissions cuts by 2030”.
The Robotics Growth Partnership, set up by the UK government, has a vision of “ensuring that smart machines play their role in … building the resilience needed to better respond to future shocks, whether that be another pandemic, climate change or something else”. Across particular sectors, robotic solutions are heralded as an important part of the quest for net zero. For example, with the expansion of wind/wave energy, the need for robot maintenance is described in “The Economic Opportunity for Robotics in Offshore Wind and Key Energy Markets”. In agri-tech, “robotics and automation create opportunities which allow for more efficient resource management and contribute to the development of more meaningful jobs” and even robots in warehouses are seen as part of more cost-efficient logistics. In addition, Innovate UK’s Transforming Food Production Challenge is “targeting net zero by 2040, with focus areas including: robotics, artificial intelligence, autonomous growing systems, precision agriculture”.
Broader digital technologies, such as the ‘Internet of Things’ (IoT), are also highlighted as a route towards net zero. A Vodafone/WPI report, Connecting for Net Zero: addressing the climate crisis through digital technology, explains how IoT and 5G, together with AI, machine learning and other “smart” technologies can substantially reduce greenhouse gas emissions. The Royal Society, with their Digital technology and the planet: towards net zero report also promotes digital technologies, with one quote stating: “Digital technologies offer a glittering prize of nearly a third of the 50% carbon emissions reductions the UK needs in the next ten years to reach Net Zero”. And all this is a view advocated by the UK Government; for example, the UK Government’s Ten Tech Priorities include “using digital innovations to reach net zero”.
It is clear that the fields of AI, IoT, robotics, and digital technologies all agree that more AI, IoT, robotics and digital technologies is vital in order to address environmental sustainability and climate change.
While some solutions based on robotics, AI or digital technologies will likely have significant benefits, we must recognise that the production and deployment of these technologies bring their own costs. Unless developed carefully, such technologies might well end up being “part of the problem”. The hardware aspects are obvious: robots, sensors, computer chips, etc, all require non-trivial resources to produce, both in terms of minerals and energy. Consequently it is vital to only use these technologies where they really will have significant benefits, far beyond the environmental costs of their production, deployment, maintenance and decommissioning. The software aspects are often under-appreciated: it may well be that the over-use of digital solutions becomes an issue in itself.
Not only does software require energy to run, but carelessly developed software might require vast communication and data storage (for example to a distant cloud server). The transfer and storage of such data may be environmentally significant, especially for ‘smart’ technologies such as AI. All of this calls for a move from simply engineering the systems that might impact upon net zero and climate change to “sustainable engineering” that takes in to account both the benefits and drawbacks of using such technologies. For software aspects this would include: devising and using efficient algorithms, in terms of energy, time, and data footprint; minimising data transfer and communications; where possible, software self-repair and reconfigurability, rather than reinstallation; minimising cloud-based and distributed activities; etc.
Manchester leading the way
In 2016, researchers now at The University of Manchester helped produce BS8611, a Guide to the Ethical Design and Application of robots and robotic systems, within the British Standards Institution (BSI).
Now, as part of its robotics activities, a committee of the BSI – led by a researcher from The University of Manchester – is producing a Guide to the Sustainable Design and Application of Robotic Systems that will include guidelines for the inherently sustainable design, protective measures and information for design and use of robots in which there is potential for environmental or ecological harm. Within this, the guide will address relevant areas in robotics, specifically the sustainability of;
- Production and fabrication
- Deployment and placement
- Energy/communication during normal/expected activity
- Failures and maintenance
- Disposal and decommissioning
- Recycling and reconfiguration.
- Clean-up and long-term effects
We need at least a re-iteration, but quite possibly a revision, of the concept of responsible innovation, wherein the environmental costs of technological `advances’ are both central and transparent. This will help us move from just the engineering of solutions to sustainability issues, and on to incorporating responsible, sustainable and ‘green’ engineering processes. Although we have highlighted robotics above, we need to expose the environmental costs of all of advanced technologies such as AI and digital, not just robotics.
The step from just engineering for sustainability, to sustainable engineering, will not only ensure that engineering is more relevant to, and aware of, environmental sustainability issues, but has the potential to drive R&D in developing less ecologically harmful artefacts, regardless of the perceived `benefits’ elsewhere.