Displacement Monitoring
A methodological framework that uses AI techniques to track and estimate the scale, direction and dynamics of large-scale population displacement caused by conflict or disaster.
Timeline (last 90 days)
AI has been applied to tracking displaced populations; by analyzing multi-source information such as changes in temporary camps in satellite images, nighttime light data, and mobile network signals, machine learning models can estimate the scale and direction of population movement.
Nighttime light data has weak signals in areas with sparse populations or unstable electricity supply, and usually needs to be cross-validated with mobile network signaling data and tent/temporary building identification results in satellite imagery to form relatively reliable estimates.
In displacement monitoring, algorithmic bias may cause certain marginalized groups (such as displaced people who have not entered formal camps and live in informal settlements) to be completely invisible to the algorithm.
All Facts (3)
AI has been applied to tracking displaced populations; by analyzing multi-source information such as changes in temporary camps in satellite images, nighttime light data, and mobile network signals, machine learning models can estimate the scale and direction of population movement.
50%UnverifiedNighttime light data has weak signals in areas with sparse populations or unstable electricity supply, and usually needs to be cross-validated with mobile network signaling data and tent/temporary building identification results in satellite imagery to form relatively reliable estimates.
50%UnverifiedIn displacement monitoring, algorithmic bias may cause certain marginalized groups (such as displaced people who have not entered formal camps and live in informal settlements) to be completely invisible to the algorithm.
50%