Skip to main content
To KTH's start page To KTH's start page

An addition to an analytical ecosystem needed to empower the lives of a billion people

Published Jan 31, 2019

Facebook releases remote sensing algorithm for approximating MV-lines

Energy is critical for development. Without it there is no development. A mother will not have access to refrigerated anesthetics for childbirth. The baby will not have access to refrigerated vaccines. As a schoolchild they will not access ICT and be initially locked out of the digital economy. Home will be filled with smoke from fuelwood burning that will have taken the mother hours to collect every day. (More than a million people a year die due to indoor smoke, and hundreds of billions of hours are lost to fuelwood collection every day). The scholar will have to be more determined than others with electricity, to study and enter university.

Fridges, electric stoves, lights all need electricity. Over a billion people do not have access to it.

Further developing investment grade electrification plans is difficult. At universities, as well as with government analysts key elements include data, software, models and support for energy planning. These might be expensive or unavailable.

Adopting open source elements of energy planning can allow for greater compatibility, transparency and efficacy. These elements - and the processes that digest and produce them - work together in some ways like an ecosystem.

The Facebook Medium Voltage line (MV-line)  data and the model that underpins it contributes to the energy planning ecosystem in important ways.

By way of background, the optimal choice of electrification technology is a function of the distance of the home to be electrified from the grid, local resource and technology availability as well as other factors. Grid information is essential.

The model developed by Facebook provides much needed advance:

It is open. Both the data that is produced and the model that underpins it. This allows for easy movement between platforms and eliminates costs. The manner in which it is documented allows the exercise to be repeated, reconstructed and audited. All are critical for developing trustworthy electrification policy and investment information.

It is a missing piece of the puzzle. There are many elements to the energy planning ecosystem. Each important. Country level grid data, and MV lines are no exception. That data is often not available, outdated or difficult to retrieve. Maps of grid lines might exist, but are not digitised. If they are digitised they might only be available as PDF files, or another format that is not easily absorbed. The new Facebook model is easily retrievable and in a standard format for ready adoption.

It adds impetus to advancing remote sensing analytics. More remote sensing data is becoming available. Advanced analytics are becoming more accessible. By providing code and results allows a growing community to scale the results to other locations and build on (or inspire new) analysis.

Further by developing and releasing this information in community the chances that a critical mass of focused advances are possible. Advances include producing new data, applying methods to new settings and by building on what is there, supporting methodological advance.

In the words of what is often quoted as an African proverb: “To go fast, go alone. To go far, go together.” Facebook and the team of collaborators who helped develop, or will be early adopters, of the model (including World Bank, KTH, WRI, University of Massachusetts) are one of several communities taking on the challenge of empowering those that are trapped, to escape poverty.