W. H. Finlay, PhD
Edmonton Regional Search and Rescue Association
Edmonton, Alberta, Canada
Email: warren.finlay@ualberta.ca
https://doi.org/10.61618/PEXK9532
Abstract
Land search and rescue (LSAR) operations often require decisions to be made between competing search strategies. Basing such decisions solely on the probability of success does not allow consideration of differences in search effort (i.e. searcher hours expended) between competing strategies. In the present work, we rely on existing optimal sequential Bayesian theory to reemphasize the utility of using a measure that considers both probability of success and effort, which we refer to as a priority rating (PR). When choosing between two competing search strategies, the more optimal strategy has a higher PR. For a search strategy employing searchers with given sweep width and travel speed in a search segment with probability density pden, PR is known to be given by PR=pdensweep widthspeed. PR values are compared for competing search strategies in a variety of scenarios, demonstrating the utility of this approach when it is used with recent developments in the literature. For example, it is found that searching by sound for a lost person in further out areas is more optimal than visual searching closer in, but this is only true out to a certain radius from the initial planning point; and dense vegetation areas have dramatically lower priority ratings, effectively concentrating search effort in them unless countermeasures are taken. The present demonstration of the use of the Bayesian priority rating approach may be helpful for search managers wanting to prioritize search efforts more effectively, including choosing the sequence to search different search segments and which search tactics are more optimal in each of these areas.
KEY WORDS: search theory, optimal, Bayesian, sequential, strategy