Where can AI infrastructure be built?
Click any parcel in France. Prometheus Search reads the land, the local plan, the high-voltage grid, water and risks, and tells you in plain words what kind of data centre could stand there, and what stands in the way.
Before you click, the territory.
- High-voltage grid
- Lines at 225 and 400 kV, and every substation from 63 kV.
- Power plants
- Nuclear, hydro, gas and renewable plants, with their capacity.
- Protected nature
- Natura 2000 sites, drawn where they block a project.
- Water
- Areas of chronic water deficit, where cooling water is scrutinised.
- Regions and departments
- Grid density and generation, territory by territory.
- Parcels
- The official cadastral plots, from street level.
After you click, the finding.
- A verdict
- No blocking constraint, points to check, or a major constraint, always with the reasons.
- What can be built
- Edge site, enterprise, hyperscale or AI campus: which fits, and why not the others.
- Next steps
- Who to call and what to ask: the town hall, the grid operator, the regulator.
- Every source
- Each line names its source, licence and date. Nothing is a black box.
Built for the people who decide where.
Developers screen dozens of sites in an afternoon, and walk into the town hall knowing the plan, the grid and the risks.
Investors check that a project's site can actually be powered, permitted and cooled before they commit.
Local authorities see what their territory can offer, and what is missing, to attract digital infrastructure.
New to planning rules or the grid? Read the guide: it explains every term in plain words.
France first. Then the rest of Western Europe.
- FranceLive
- LuxembourgIn preparation
- BelgiumIn preparation
- NetherlandsIn preparation
- GermanyIn preparation
Help shape Prometheus Search.
Request access to the map, then tell us what helped and what needs work. Your feedback will guide what we build next.
What would make this useful in your work? Share the place you explored, what you wanted to find out, and anything that was missing or unclear.
Share feedback on LinkedIn