What are the specific barriers for hedgehogs to find food and mates?
For decades, Britain's hedgehogs have been quietly disappearing, their decline a symptom of a landscape increasingly hostile to small, wandering creatures. Now, researchers at Cambridge University are turning to satellites and artificial intelligence to read that landscape with a precision no human eye could achieve alone — mapping hedgerows, predicting habitats, and following individual animals through the night to understand not just that hedgehogs are vanishing, but exactly where and why the world has become impassable for them. It is a story about what it means to truly see a crisis, and whether seeing it clearly enough, in time, can change its course.
- Hedgehog populations across Europe have collapsed over recent decades, driven by habitat loss, agricultural intensification, and a countryside increasingly fragmented by roads and development.
- Cambridge's AI system, Tessera, trained on roughly 20 petabytes of satellite imagery, can identify individual hedgerows and predict suitable habitats even through cloud cover — offering a level of landscape detail that traditional field surveys cannot match.
- Some hedgehogs have been fitted with GPS trackers, their nightly journeys overlaid onto satellite maps to reveal the invisible barriers blocking their access to food, mates, and safe passage.
- The open-source Tessera platform has already been adopted by over 100 research groups, extending its reach into farmland monitoring and agricultural tracking across the UK.
- A genuine tension runs through the project: critics warn that the energy demands of training and running powerful AI systems may themselves contribute to the environmental pressures threatening the very species researchers are trying to save.
Britain's hedgehog population has been in freefall for decades, and researchers at Cambridge University are now combining two unlikely tools — orbiting satellites and artificial intelligence — to understand why. At the heart of the effort is Tessera, an AI system trained on approximately 20 petabytes of satellite imagery that can read the British landscape in remarkable detail: identifying individual hedgerows, mapping habitat patches, and predicting where hedgehogs might live even in areas hidden beneath cloud cover.
What makes the project distinctive is its pairing of this aerial view with ground-level tracking. Some hedgehogs have been fitted with tiny GPS devices, their real-time movements overlaid onto the satellite maps to reveal patterns invisible to conventional surveys. A single animal's nightly journey across a mapped landscape can expose whether it is moving through viable habitat or struggling against barriers that conservation efforts might actually address.
Building Tessera demanded extraordinary effort. Cambridge researchers initially exhausted their allocated computing power, installing additional processors under their desks before a partnership with technology firms AMD and Vultr allowed the work to scale. The system is now open-source, accessed by more than 100 research groups for purposes ranging from hedgehog conservation to monitoring crop rotations across UK farmland.
The project is not without its tensions. Some conservationists have raised pointed questions about the environmental cost of running energy-intensive AI systems — arguing that the technology's own footprint may contribute to the pressures threatening species like hedgehogs. The researchers acknowledge the concern but press forward, reasoning that continuing to lose hedgehogs without understanding why carries its own cost. What emerges is a portrait of modern conservation: satellites, algorithms, and small animals carrying GPS trackers, all working together to decode the quiet disappearance of a creature that once thrived across the British countryside.
Britain's hedgehog population has been collapsing for decades, and researchers at Cambridge University are now deploying an unlikely pair of tools to understand why: satellites orbiting overhead and artificial intelligence running on the ground. The effort centers on Tessera, an AI system trained on roughly 20 petabytes of satellite imagery—equivalent to about 10 billion standard digital photos—that can read the British landscape in extraordinary detail. From space, it identifies individual hedgerows, maps habitat patches, and even predicts where hedgehogs might live in areas obscured by cloud cover. The goal is not just to see where these small mammals survive, but to pinpoint the specific obstacles keeping them from finding food, locating mates, and moving safely across the countryside.
The scale of the hedgehog crisis is stark. Populations across Europe have declined dramatically in recent decades, a collapse driven by habitat loss, fragmentation, and the intensification of agriculture. What makes the Cambridge project distinctive is its attempt to combine satellite surveillance with ground-level tracking. Some hedgehogs have been fitted with tiny GPS devices—researchers call them "digi-hogs"—that transmit their real-time movements. When overlaid with the satellite maps showing which landscapes are hedgehog-friendly and which are becoming hostile, the data begins to reveal patterns invisible to traditional field surveys. A single hedgehog's nightly journey across a mapped landscape can show whether it's navigating through suitable habitat or struggling to cross barriers that shouldn't be there.
Professor Silviu Petrovan, strategy and research manager at People's Trust for Endangered Species, frames the work with cautious optimism. The hope is that these powerful computational models can isolate the precise factors limiting hedgehog survival—not just broad categories like "habitat loss," but specific, actionable barriers that conservation efforts might actually address. The satellite data also allows researchers to track how new housing developments, road construction, and other environmental changes reshape the landscape over time, creating a historical record of how human activity has carved up the spaces where hedgehogs live.
Building Tessera required extraordinary computational effort. The researchers at Cambridge initially exhausted their allocated computing power and began installing additional processors under their desks to continue training the system. A partnership with US technology firms AMD and Vultr eventually provided access to more infrastructure, allowing the work to scale. The system itself is open-source, meaning it's not locked away for Cambridge's exclusive use. More than 100 research groups have already accessed it, adapting it for other purposes: monitoring farmland, tracking which crops are grown in which fields across seasons, building a detailed picture of UK agriculture.
Professor Anil Madhavapeddy, who leads planetary computing at Cambridge, explains that satellite imagery is inherently messy. Clouds obscure the view, lighting changes between day and night, and the raw data requires extensive processing before it becomes useful. Tessera compresses all that complexity into clean, easy-to-read maps that researchers can interrogate with specific questions. The system answers queries that would be nearly impossible to pursue manually: Where are hedgehogs most vulnerable? Which habitat patches are becoming isolated? What does a landscape need to support a healthy population?
The project does not exist without controversy. Some conservation advocates have raised concerns about the environmental cost of training and running power-hungry AI systems. The energy demands of machine learning, they argue, may themselves contribute to the climate and environmental pressures that threaten species like hedgehogs. It's a tension the researchers acknowledge but believe is worth navigating. The alternative—continuing to lose hedgehogs without understanding why—seems worse. What emerges is a portrait of modern conservation: not boots on the ground alone, but satellites, algorithms, and GPS trackers working in concert to decode why a creature that once thrived across the British countryside is now vanishing.
Notable Quotes
We're hoping to understand what are the very specific barriers for hedgehogs to find food and find their mates, and be able to safely move around the countryside.— Prof Silviu Petrovan, People's Trust for Endangered Species
Tessera compresses loads of satellite data and gives us really easy-to-use maps of the UK, where we can ask really specific questions about things we can see from space.— Prof Anil Madhavapeddy, Cambridge University
The Hearth Conversation Another angle on the story
Why does it matter that we can see individual hedgerows from space? Couldn't researchers just walk around and count hedgehogs?
Walking finds the hedgehogs that are there. Satellites show you the entire landscape at once—every field, every gap, every place a hedgehog could theoretically live but doesn't. That's how you spot the barriers.
What kind of barriers are we talking about?
A road with no crossing points. A field sprayed with pesticides so there's nothing to eat. A housing estate that carved a habitat in half. The GPS trackers on individual hedgehogs show the problem; the satellite maps show the scale of it.
So the "digi-hogs" are the ground truth?
Exactly. One hedgehog's nightly route tells you whether it's finding food, whether it's safe, whether it's isolated from potential mates. Multiply that across dozens of tracked animals and you start seeing patterns the satellite data alone couldn't reveal.
Why does it take 20 petabytes of training data to teach an AI to recognize a hedgehog habitat?
Because the AI isn't just learning what a hedgehog looks like. It's learning to read the entire landscape—vegetation patterns, soil moisture, field boundaries, the texture of places where hedgehogs can survive. That requires seeing millions of examples.
And the environmental cost of all that computing?
That's the real tension. You're burning energy to save an animal threatened partly by the energy-intensive world we've built. But the researchers argue that without understanding the problem precisely, conservation efforts are just guessing.
What happens next?
The maps get shared with conservation groups and local planners. They show where to restore habitat, where to create crossing points, where development should be restricted. The hope is that understanding the problem at this level of detail actually makes solutions possible.