Facebook, as part of their internet.org initiative, is working on bringing internet access to people in rural areas all across the world. For obvious socioeconomic reasons, this is an important initiative. From a business standpoint, it also grows the base of potential Facebook users at a time where that top line number is starting to plateau.

In order to understand how people are settling and aggregating throughout the world, Facebook has been using high-resolution population maps and then creating models to drill down and analyze individual buildings. 

This is interesting because: how else could you track informal settlements? So far this seems to be the most effective method. In fact, in architecture school I worked on a project in Dhaka, Bangladesh and I remember relying heavily on aerial photography because I simply couldn’t find the data I was looking for.

Here’s an excerpt from a recent World Bank post. They, along with Columbia University, are collaborating with Facebook.

Facebook’s computer vision approach is a very fast method to produce spatially-explicit country-wide population estimates. Using their method, Facebook successfully generated at-scale, high-resolution insights on the distribution of buildings, unmatched by any other remote sensing effort to date. These maps demonstrate the value of artificial intelligence for filling data gaps and creating new datasets, and they could provide a promising complement to household surveys and censuses.

And here’s an excerpt from a recent Facebook post on the same topic:

From this preliminary analysis, we’ve determined that slightly less than 50% of the population lives in cities. However, 99% of the population lives within 63 km of the nearest city. Hence, if we are able to develop communication technologies that can bridge 63 km with sufficiently high data rates, we should be able to connect 99% of the population in these 23 countries. [”These 23 countries” represents about 1/3 of the world’s population.]

I would imagine that this type of model works better in sparsely populated / low-rise areas. Still, I am very interested in thinking about ways in which our current survey and census methods could be improved. Here in Canada we conduct our national census every 5 years. In today’s world that feels like eons. One day, I am sure, it will be real-time.

Image: World Bank

Facebook, as part of their internet.org initiative, is working on bringing internet access to people in rural areas all across the world. For obvious socioeconomic reasons, this is an important initiative. From a business standpoint, it also grows the base of potential Facebook users at a time where that top line number is starting to plateau.

In order to understand how people are settling and aggregating throughout the world, Facebook has been using high-resolution population maps and then creating models to drill down and analyze individual buildings. 

This is interesting because: how else could you track informal settlements? So far this seems to be the most effective method. In fact, in architecture school I worked on a project in Dhaka, Bangladesh and I remember relying heavily on aerial photography because I simply couldn’t find the data I was looking for.

Here’s an excerpt from a recent World Bank post. They, along with Columbia University, are collaborating with Facebook.

Facebook’s computer vision approach is a very fast method to produce spatially-explicit country-wide population estimates. Using their method, Facebook successfully generated at-scale, high-resolution insights on the distribution of buildings, unmatched by any other remote sensing effort to date. These maps demonstrate the value of artificial intelligence for filling data gaps and creating new datasets, and they could provide a promising complement to household surveys and censuses.

And here’s an excerpt from a recent Facebook post on the same topic:

From this preliminary analysis, we’ve determined that slightly less than 50% of the population lives in cities. However, 99% of the population lives within 63 km of the nearest city. Hence, if we are able to develop communication technologies that can bridge 63 km with sufficiently high data rates, we should be able to connect 99% of the population in these 23 countries. [”These 23 countries” represents about 1/3 of the world’s population.]

I would imagine that this type of model works better in sparsely populated / low-rise areas. Still, I am very interested in thinking about ways in which our current survey and census methods could be improved. Here in Canada we conduct our national census every 5 years. In today’s world that feels like eons. One day, I am sure, it will be real-time.

Image: World Bank