How artificial intelligence can turn climate goals into sustainable reality
Squeezed by environmental concerns, organisations and companies are turning to machine learning to limit their impact on the climate
Tracking and reducing a company’s carbon footprint can be a bewildering task; the largest firms have multiple operations and data streams, and often overlapping functions. This complexity is one reason why some companies, even the most committed and ambitious, are struggling in the race against climate change.
Artificial intelligence and advanced analytics can help. AI can cut through complicated data and help organisations speed up their actions on climate change. It can help in the mitigation of emissions and support adaptation to the impacts of climate change by calculating physical risks, measuring cost of inaction and planning adaptation strategies.
A global survey of more than 3,000 executives revealed that more than
of respondents are deploying AI.
had an AI strategy in 2020
From measuring and reducing emissions at scale to enabling innovative business models, AI can help turn climate goals into sustainable reality. And, because AI can analyse large and unstructured data sets, it can help identify insights from otherwise challenging sources.
1. Specific examples of where AI can help include:
Use the carousel to explore
“When applied responsibly, AI can be a powerful tool in the fight against climate change because its use opens up many new avenues to quickly analyse information, make informed decisions, and move to action,” says Hamid Maher, Managing Director and Partner at Boston Consulting Group (BCG). “AI can help tackle the most complex challenges impeding large-scale reduction of carbon emissions, as well as adaptation and resilience efforts.”
BCG estimates that using AI across industry could achieve
of the carbon reduction needed to limit warming to
between 2.6 and 5.3 gigatons of CO2e.
2. Power of patterns
The great strength of AI lies in its capacity to learn by experience, collecting massive amounts of data from its environment, finding patterns and connections that people fail to notice, and recommending appropriate actions.
For example, BCG recently worked with an oil and gas company facing production losses due to unexpected difficulties with machinery, and a control system that could only react to changes after they happened. To compensate for the losses, the company had to increase production, leading to higher emissions and increased costs. AI helped to rectify the problem. Introducing machine learning to monitor the performance of plant equipment allowed a new control strategy based on making predictions about future performance and then acting to improve it.
The company’s new AI system can predict maintenance problems and CO2 emissions for each production unit, giving plant engineers the information they need to predict energy use and emissions for multiple units several hours into the future. That means they can isolate and fix any equipment producing excess emissions.
As a result, the oil & gas company lowered its carbon emissions by up to
of greenhouse gases per year, and reduced its costs by some $5mn to $10mn.
“If we can scale up these kinds of gains across the wider industry and other sectors, we could make a real and rapid difference,” says Mateo Garcia-Novelli, Consultant at BCG. “Company leaders have a clear opportunity to use AI to target areas with high carbon emissions and significant costs, especially those with a potential payback period of less than two years.”
Companies that have scaled AI across the business and achieved meaningful value from their investments typically dedicate:
3. Great Interest
Companies are certainly interested in the opportunities offered by advanced analytics tools and machine learning. For a recent AI for the Planet Alliance report, BCG asked global public- and private-sector climate and AI leaders about the potential of AI as a tool in the fight against climate change.
But the survey also revealed significant obstacles. Some 78 per cent of those surveyed cite low access to AI expertise, whether inside or outside their organisation, as an obstacle, while 77 per cent report a lack of available AI solutions. And 67 per cent say they face lack of organisational confidence in AI data and analysis.
Garcia-Novelli says there are some clear steps that could help more organisations access and exploit AI-based routes to greater sustainability. The first is to make them as user-friendly and accessible as possible, irrespective of whether they are designed for companies, governments or the public. Second, AI solutions need dedicated support to bridge the gaps between academic research, proof of concept and at-scale deployment. Such support could include access to mentorship and capital, as well as training and reskilling to ensure decision-makers can understand and apply AI tools.
Today’s AI research is predominantly run out of institutes and corporations in the global north. Resources must be directed towards ensuring the development of AI technologies by and for the global south, where many countries will be disproportionately affected by the changing climate.
AI can gather and analyse large datasets in real time, leading to the creation of early warning systems for extreme weather events, and long-term projections of localised events such as sea-level rise.
This will be especially vital for the more than 3bn people living in areas highly vulnerable to climate change, particularly in the global south.
“AI cannot solve climate change on its own but it opens up so many new avenues of analysis, helping leaders make informed decisions faster on a topic where action is urgently needed,” says Maher. “We know it can work. Now we need to make it work at scale.”