How have researchers from Peking University and Alibaba’s Damo Academy used AI and satellite imagery to create a nationwide inventory of China’s renewable energy infrastructure, including mapping hundreds of thousands of solar photovoltaic installations and wind turbines from 7.5

• How they built it

  • The team combined AI-based image recognition with large-scale satellite remote sensing to detect and locate renewable-energy assets across the country rather than relying only on fragmented administrative records or modeled estimates.
  • DAMO’s models were run in the cloud to handle the huge image volume and China’s varied terrain, which the reports say was a key technical hurdle in mapping facilities nationwide at high precision.
  • The resulting dataset was reported as being published in Nature on May 20, 2026.

• What the inventory contains

  • The map reportedly covers solar PV sites and wind turbines nationwide, with counts of about 319,000 PV facilities and 91,600 wind turbines.
  • It spans 1,915 counties, which suggests a much more spatially complete picture of where renewable capacity is actually located.

• Why the “God’s-eye view” matters

  • A national, geolocated inventory can help grid planners optimize transmission, dispatch and balancing because they can see where wind and solar assets are concentrated and how those patterns vary across regions.
  • The researchers and follow-up coverage say it can also support environmental assessment and broader energy-system analysis, including identifying where wind and solar complement each other and where curtailment risks may be higher.

• Why this matters for data-centre demand

  • China is also trying to connect renewable power more directly to fast-growing computing loads: one recent project in Ningxia was described as the country’s first large-scale scheme to supply renewable energy directly to data centres.
  • Coverage of the mapping project explicitly links the inventory to rising electricity demand from data centres and AI computing, arguing that a clearer map of wind and solar resources could help stabilize the grid as those loads expand.
  • In practice, this kind of map could help China decide where to site new data centres, where to add transmission or storage, and where renewable generation can best meet computing demand with lower carbon emissions.

If you want, I can also turn this into a 5-sentence executive summary or explain the likely AI/computer-vision pipeline behind the mapping.