Britain’s hedgehog problem is not a lack of affection for the animal. It is a lack of precise, up-to-date visibility into where the species can still move, feed and breed.
Researchers at the University of Cambridge are using satellite data and AI to map hedgehog habitats across the UK, in a project aimed at slowing a steep population decline, according to BBC Tech. The tool at the center of the work is Tessera, an open-source AI system that analyzes detailed satellite imagery and turns it into maps that can show hedgerows, habitat loss and likely hedgehog-friendly areas even when cloud cover obscures the view.
The tension is obvious. Conservation has long depended on field observations, local records and targeted surveys. Cambridge is now adding something more industrial: 20 petabytes of training data, extra compute from AMD and Vultr, and satellite maps detailed enough to track change across landscapes over time.
Britain expected surveys to carry the load; Cambridge is adding satellites
The urgency comes from the scale of the decline. Hedgehog populations have fallen sharply across Europe in recent decades. In the UK, a 2022 report estimated numbers had dropped by up to 75% in rural areas since 2000, according to the BBC report. The common western European hedgehog, described by the RSPCA as the UK’s only native species, has also been listed as “Near Threatened” by the International Union for the Conservation of Nature.
That does not tell conservationists enough on its own. A national decline figure can show the severity of the problem, but it cannot show which hedgerow has disappeared, which route has been cut off, or which development has changed a local habitat network.
That is where Tessera changes the workflow. The system analyzes images of the UK gathered from space and produces maps that capture landscapes in fine detail, including individual hedgerows. The goal is not just to show where hedgehogs may live. It is to identify where habitat is vanishing and where barriers may prevent hedgehogs from finding food and mates.
Prof Silviu Petrovan, strategy and research manager at People's Trust for Endangered Species, framed the core question this way:
“What we're really hoping is that we can use these really powerful models to understand, for instance, what are the very specific barriers for hedgehogs to find food and find their mates, and be able to safely move around the countryside?”
That is the practical value. The project is less about producing a prettier wildlife map and more about giving conservationists a sharper way to decide where to look, where to measure, and where changes in the landscape may be doing damage.
The map is about habitat, not hedgehogs visible from orbit
A key point can get lost in the headline: researchers are not using satellites to count individual hedgehogs from space. The reported use case is habitat mapping.
Tessera identifies landscape features that matter to the project, including hedgerows and other terrain patterns visible in satellite imagery. It can also predict hedgehog-friendly areas that are hidden by cloud cover. That matters because satellite data is messy, inconsistent and often obstructed.
Anil Madhavapeddy, professor of planetary computing at Cambridge University, described the problem bluntly:
“Satellite data is really complicated to use and really noisy, because you have to do things like cloud removal and adjust for day and night, and so on.”
Tessera’s job is to compress and clean that complexity into usable maps. In Madhavapeddy’s words:
“Tessera compresses loads of that 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.”
That distinction matters. A habitat map can reveal the conditions around hedgehogs. It can show where hedgerows are present, where they disappear, and how environmental change may reshape the places hedgehogs can use. It does not, by itself, prove where every animal is.
The stronger model is combined evidence. Cambridge’s maps can be paired with other data sources, including tiny GPS trackers physically attached to some hedgehogs. Researchers involved in the project refer to hedgehogs with trackers as “digi-hogs.” A similar tracker initiative is already under way in Northern Ireland, according to the BBC.
Tessera turns noisy satellite data into conservation questions
The AI component matters because the dataset is too large and too noisy for manual inspection at national scale. Tessera was trained on around 20 petabytes of data, described by the BBC as equivalent to 10 billion standard digital photos.
That scale created a compute problem inside Cambridge. Researchers hit the limits of the university’s allocated computing power and installed extra processors under their desks to keep the work going. A new deal with AMD and Vultr has since given the project access to additional infrastructure.
The before-and-after shift is straightforward:
- Before: Conservation teams relied heavily on fieldwork, local observations and narrower datasets to infer where hedgehog habitat was changing.
- After: Researchers can use AI-assisted satellite maps to scan landscape change at much larger scale, then decide where field validation or tracker data should be focused.
- Before: Cloud cover and satellite noise made raw imagery difficult to translate into practical conservation maps.
- After: Tessera can process that noise and produce easier-to-use maps for specific questions about visible landscape features.
- Before: A housing development or environmental change might be assessed locally.
- After: Researchers say Tessera outputs can help track the impact of new housing developments and other environmental changes over time.
The system is not limited to hedgehogs. It is open source, and Madhavapeddy told the BBC that more than 100 research groups have accessed it. Tessera can also monitor farmland and track which crops are being grown in which fields over time, building a detailed picture of UK agriculture.
That broader use matters for the hedgehog project. The same mapping engine that classifies changes in farmland can also help researchers understand how rural habitat structure is shifting. Analysis: this is where the project becomes more than a species-specific tool. It is a test of whether general-purpose environmental AI can produce conservation-grade evidence.
A local hedgehog map would change the search pattern first
The most useful near-term output is not an instant fix. It is prioritization.
Imagine a local area where hedgehogs are recorded in one set of places but not in nearby areas that appear, from the ground, to be reachable. The BBC source does not specify a real neighbourhood case study, and it does not list the exact local barriers Tessera has already found. But the reported method shows how conservationists could investigate that kind of problem.
A Tessera-assisted workflow would start with the map:
- Locate habitat features: Identify hedgerows and other mapped landscape elements associated with hedgehog-friendly places.
- Compare change over time: Check whether new housing developments or other environmental changes have altered those features.
- Add movement data: Combine the satellite output with GPS tracker data from “digi-hogs” where available.
- Test the prediction: Use fieldwork or local records to confirm whether the AI map reflects real hedgehog movement and habitat use.
That is a different conservation posture. Rather than spreading effort evenly across a broad area, researchers can use the model to narrow the search. The map says: this is where the landscape changed; this is where habitat may be fragmented; this is where tracker data or field checks could be most informative.
For councils, landowners and conservation groups, the value would be operational. Analysis: if the maps are accurate, they could help turn vague concern about hedgehog decline into a more targeted list of sites that deserve attention. The BBC report supports that direction but does not yet show evidence of specific interventions succeeding because of Tessera.
The model has blind spots: cloud, compute and proof on the ground
The project also carries limits.
First, satellite imagery sees the landscape from above, not the full ecological reality on the ground. Tessera can map features visible from space and infer hedgehog-friendly areas, but the BBC report does not say it can capture every small-scale condition that may affect an individual animal’s movement or survival.
Second, AI predictions still need validation. A model can flag patterns. It cannot replace ecological expertise, tracker data, field surveys or local evidence. That is especially important when the output may influence conservation priorities or planning decisions.
Third, the compute burden is not trivial. The system was trained on 20 petabytes of data, and Cambridge researchers needed more processing power after exhausting their university allocation. The BBC also notes that some observers have urged caution around power-hungry AI because of environmental concerns. For a conservation project, that trade-off cannot be waved away.
The credibility test is therefore not whether AI can produce impressive maps. It can. The test is whether those maps produce better decisions than existing methods, at a cost and energy profile that conservationists can justify.
The next break point is planning, not prediction
The promise of AI hedgehog tracking is realistic if it stays narrow: better maps, better questions, better targeting of scarce conservation work.
Tessera can help researchers locate habitats, detect where they are disappearing, and examine how new housing developments or other environmental changes affect landscapes over time. Combined with GPS data from “digi-hogs,” it could help reveal the barriers that stop hedgehogs from reaching food and mates.
But satellites and AI will not reverse the UK hedgehog decline by themselves. They are measurement tools. The real-world impact depends on what happens after the map is generated: whether researchers validate the signals, whether conservation groups act on them, and whether planning decisions account for the habitat changes the system exposes.
The watch item is simple: Cambridge’s project now needs to move from sharper visibility to demonstrated conservation outcomes. If Tessera can show not only where hedgehog habitat is changing but where intervention should be prioritized, AI will have earned a practical role in one of Britain’s more visible wildlife declines.
Impact Analysis
- UK rural hedgehog populations have dropped by up to 75% since 2000, making faster habitat insight urgent.
- AI and satellite mapping could help conservationists identify where hedgehogs can still move, feed and breed.
- The project shows how large-scale computing and open-source AI are becoming tools for wildlife protection.
