Navigation of UAVs for tracking of atmospheric release of radiation

Navigation of unmanned aerial vehicles (UAVs) is a popular and well studied topic. In this contribution, we apply the existing decision theory to the problem of tracking atmospheric pollution, specifically accidental release of radiation. The UAVs work in tandem with the stationary radiation monitoring network to estimate consequences of the radiation accident. Results of this estimation serve as supporting material for the human decision makers, whose main concern is protection of the population. We test navigation of the UAVs with respect to two loss functions: mutual information and expected misclassification of affected people. Several approximations were designed to navigate the UAVs in real time. The simulation experiments confirm that UAVs have great potential to complement limited capabilities of the stationary radiation monitoring network.