Rescued at sea by algorithms“The helicopter pilots who found John didn’t think they found John because they were particularly awesome rescue pilots. Instead, they said, they found him because of the whole high-tech search-and-rescue apparatus that the Coast Guard is able to deploy in cases like this. As much as anything, they were proud of their algorithms.”— NYT reporting on the rescue of fisherman John Aldridge
The US Coast Guard Search and Rescue Optimal Planning System SAROPS includes a drift and rescue simulator to compute the possible distress positions and times, subsequent search object drift trajectories, and the effect of completed searches on the search object probabilities.
The simulator uses algorithms classed as Monte Carlo Methods that are generally used for simulating systems with many degrees of freedom such as fluids, and also for predicting systems with uncertainty in inputs such as business risk. SAROPS can simulate drifts for up to 10,000 particles and accounts for changes in water current, wind leeway and leeway divergence. The results are displaced on an animated density map.
Looks like a GUI implementation of search processes discussed in The Theory that Would Not Die, a very readable intellectual history of Bayes’ Theorem. Cool!