Researchers from the University of Cambridge have shared how artificial intelligence (AI) is helping tackle some of the world’s most urgent environmental problems.
Image Credit: U.P.SD/Shutterstock.com
The study highlights four key initiatives: Terra, a platform for modeling terrestrial ecosystems; AI-assisted synthesis of conservation evidence; machine learning to improve climate modeling; and thermal imaging for enhancing housing efficiency.
Together, these projects reflect how AI can support biodiversity protection, improve climate forecasts, and promote more sustainable urban planning. By combining large-scale data with deep scientific knowledge, researchers aim to inform smarter decisions around land use, conservation, and energy, balancing human needs with the health of the planet.
The Urgency Behind the Research
Traditional approaches are increasingly overwhelmed by the scale and complexity of the planetary crisis. With over 60 % of global ecosystems degraded, growing uncertainty in climate models, and cities facing mounting energy inefficiencies, innovative solutions are critical.
Cambridge researchers from a range of disciplines are responding with AI-powered tools that can process environmental data at an unprecedented scale. Professor Anil Madhavapeddy and his team are developing Terra to model terrestrial ecosystems. Professor Bill Sutherland is using AI to synthesize decades of conservation evidence. Climate scientists are applying machine learning to refine predictions, while architects are deploying thermal imaging algorithms to pinpoint energy inefficiencies in housing.
Despite their different domains, these projects face common challenges: integrating diverse datasets, speeding up evidence-based decisions, and creating interventions that work across local and global scales. What unites them is a shared goal—bridging technical innovation with environmental science to build practical, actionable tools that guide better policy decisions.
From Global Models to Local Actions: Biodiversity and Conservation
At the ecosystem level, Terra represents a new approach to modeling biodiversity. By integrating satellite imagery, drone footage, and field data, the platform forecasts how different land-use scenarios could affect biodiversity. One focus is identifying “exchange zones”—areas where agricultural development could expand with the least ecological impact. Rather than pushing for simplistic fixes like full rewilding or intensive farming, the model helps map out realistic tradeoffs.
Complementing this broader view, the Conservation Evidence project, led by Professor Sutherland, has reviewed 1.6 million studies to identify which conservation strategies actually work. Using a curated database, their AI system can now analyze new research—across multiple languages—up to 100 times faster than a human. Dr. Sadiq Jaffer’s Conservation Co-Pilot builds on this work by offering tailored, actionable recommendations, such as specific grassland management techniques to protect endangered species.
These tools address a long-standing issue in conservation: the gap between published research and practical application. By validating the AI’s output against known evidence, the team ensures both speed and reliability. Importantly, both Terra and the Co-Pilot are designed to enhance, not replace, human expertise. Terra helps users navigate systemic tradeoffs, while the Co-Pilot delivers locally relevant guidance.
Smarter Climate Forecasting and More Efficient Cities
Cambridge’s climate scientists are also using AI to push the boundaries of forecasting. Dr. Jack Atkinson’s FTorch software enhances traditional climate models by improving how they simulate small-scale phenomena like cloud formation. The result was that the models ran 40 % faster without sacrificing accuracy, while also using less energy to compute.
Dr. Joe Wallwork’s team is applying similar machine learning techniques to better predict extreme weather events. These improvements could mean earlier, more accurate warnings—and potentially save lives.
In cities, the focus shifts to energy efficiency. Dr. Ronita Bardhan has developed a thermal imaging algorithm that analyzes satellite data to detect heat-loss patterns across entire urban areas. In Cambridge, the tool has already mapped 700 homes in need of retrofits, helping prioritize upgrades for maximum energy savings. It also assesses overheating risks in summer, offering a dual solution for climate adaptation.
What’s striking about both climate and urban projects is their ability to connect different scales, linking localized data (like heat escaping from a home) with broader patterns (such as national energy policy). FTorch is already being used by leading climate research institutes, and Bardhan’s work is shaping UK energy efficiency guidelines.
Looking Ahead
These initiatives show what’s possible when AI is applied thoughtfully to environmental problems. By combining large-scale analysis with local context, the tools enable more targeted and equitable responses to today’s most pressing ecological challenges.
Across all projects, a few key themes emerge: interdisciplinary collaboration, a commitment to open-access tools where possible, and the importance of keeping humans in the loop. Rather than replacing experts, AI is helping them work faster and smarter, bridging the gap between data and action.
Whether it’s helping conservationists pinpoint effective strategies or guiding homeowners on energy upgrades, these tools are making critical knowledge more accessible and actionable. As these technologies evolve, they hold real potential to support more informed decision-making during this crucial decade for planetary health.
Disclaimer: The views expressed here are those of the author expressed in their private capacity and do not necessarily represent the views of AZoM.com Limited T/A AZoNetwork the owner and operator of this website. This disclaimer forms part of the Terms and conditions of use of this website.