Gridfinder work published in Nature SciData!

Over the last year, I did some very fun work using satellite imagery (night-time lights) and OpenStreetMap data to predict the location of global electrical grid infrastructure. I wrote about it briefly on this blog here and as a World Bank blog here. And now finally its seeing the light of day as a published article in Nature Scientific Data:

Predictive mapping of the global power system using open data

You can also visualize the full global dataset at gridfinder.org and download the full dataset. Please have a look, use it if its useful, and keep in mind the (big!) caveats before treating it as ground-truth.

A screenshot from gridfinder.org showing Southern Africa.
A screenshot from gridfinder.org showing Southern Africa.