COP27: How we’re applying AI to benefit the environment
COP27 brought together major world leaders, policymakers, and climate activists in Egypt to plan for a greener future. The environmental picture can often seem bleak, with even UN Secretary-General António Guterres declaring at the summit that ‘we’re on the highway to climate Hell.’ Action is needed now and new ideas need to be brought to the table to realise the UK’s ambitions for Net-Zero status by 2030.
AI can be applied to help achieve these targets. From measuring and offsetting our own emissions to using AI to bolster the clean energy transition, we’re passionate about innovating our way to a sustainable future.
Taking environmental accountability
As the UN Secretary-General calls out ‘bogus net-zero pledges to cover up massive fossil fuel expansion’ that ‘could push our world over the climate cliff’, we’re eager to take environmental accountability.
We are a founding member of the Tech Zero taskforce, a group of tech companies with the shared view that we can make significant reductions in emissions quicker by working together. We began our efforts to reduce our emissions by working with Supercritical, a carbon measurement and offsetting specialist, to measure our emissions.
Our emissions include obvious contributors, such as our office use and employee travel, as well as less predictable causes including procurement, events, and areas linked to our work such as cloud servers. Fast growth companies such as ours scale at great pace, adding dozens of new employees each month and scaling our cloud server use. As we scale so too do our emissions and therefore we knew our carbon offsetting had to match our growth.
We invest in offsets that can permanently reduce our emissions. This is achieved through investment in biochar and bio-oil sequestration, in which the carbon is sequestered for 10,000 years.
However, beyond offsetting all of our own carbon emissions, we are also looking to apply our AI expertise to projects that yield positive environmental results.
Applying AI to the clean energy transition
Our belief in the potential of AI to address environmental issues led to us creating our Energy Transition and Environment business unit over a year ago. Through this, we are working with DRIFT Energy to apply our AI to the world’s first hydrogen generating boat.
Renewable energy often depends upon uncontrollable weather. DRIFT aims to mitigate this by operating boats capable of following the wind, using a route optimisation algorithm we have developed to navigate in the most effective way to harness maximum wind power. Rather than waiting for the right conditions, DRIFT’s vessels can go out and find them.
The environmental benefits are clear: the average UK offshore wind farm generates below 40% of their potential capacity as a result of fluctuating conditions. Initial testing suggests DRIFT’s mobile approach has the potential to achieve 70-80%.
The combination of DRIFT’s hydrogen-generating boat and our AI optimisation represents an exciting opportunity to bolster renewable energy ambitions. The aim now is to optimise boat design, refine the algorithm and move towards operational readiness. Innovative green hydrogen generation such as this illustrates how AI can enable and super-charge the global journey to Net-Zero.
Mitigating emissions early on
Applying AI to generate clean energy is one way of shoring up a green future, but it’s also key that we also minimise harmful emissions in the present. Our Energy Transition and Environment business unit has recently partnered with SGN and Utonomy to explore how advanced machine learning techniques could help optimise pressure within gas networks, contributing to reduced methane leakage.
More broadly, the partnership has developed a number of opportunity areas where artificial intelligence could be deployed with remote pressure control technology to improve decision-making, and ultimately deliver benefits to customers, the environment, and network planners.
The ambition behind the project is to build and implement proof-of-concept machine-learning models as an end-to-end solution, integrated across the network’s data architecture. In doing so, the models will augment existing pressure management decision-making and reduce avoidable leaks.
Future Data Scientists work to better our future climate
We bring our passion for the environment into many of our Fellowship projects. Our Fellowship, which trains the brightest data science talent and pairs them with host companies, has seen many fellows apply AI to environmentally advantageous causes.
In a recent programme one of our Fellows, Ned, worked with the campaigns team of the environmental charity WWF. Using machine learning models to segment audiences, this project deepened our understanding of the UK public’s relationship with food and diets in relation to environmental impact. These results will support future campaigns and policy work.
Additionally another Fellow, Faith, worked with the Fossil Fuel Non-Proliferation Treaty Initiative, a global NGO which aims to empower policy advisors and researchers in their efforts to facilitate a global transition to clean energy. This included building an interactive online app which can automatically identify, filter, and visualise relevant policies globally to help the NGO. Many more Fellows have applied their AI expertise to environmental NGOs, Government divisions, and charities with focuses on sustainability and using technology to achieve a greener future.
Paving the way to a greener future
The breadth of our projects highlights the power of AI applications to better the environment. From working to address fossil fuel emissions and mitigate leaks, to innovative new solutions able to generate clean energy, artificial intelligence is a technological solution that can play an influential role in protecting our planet’s future. While we might be ‘on the highway to climate Hell’, AI is a powerful tool to help us navigate us to a different, Net-Zero future.