Multiobjective Spatial Optimization: The Canadian Coast Guard

Abstract

This paper examines a problem of the Canadian Coast Guard. The mission of the Coast Guard (Atlantic) is to provide assistance to vessels or persons aboard that are in distress. This piece focuses on the search & rescue mission of the Coast Guard. The two most important criteria include the average time that it takes to reach a vessel in distress, and the number of vessel incidents that can be reached within a reasonable time limit. Given that demand throughout the year is not constant but peaks during the summer season (in excess of 50% of the distress calls occur in July and August), there is a potential for congestion and the resulting inability of the Coast Guard to answer distress calls within an acceptable amount of time. In order to avoid formulating the model as a probabilistic model with all its computational difficulties, we use backup coverage as a third important concern. We model the problem of locating different types of rescue vessels along the coastline of the Maritime Provinces as an integer programming problem with three objectives, each dealing with one of the major concerns. The problem is solved given the available vessels and on the basis of incidents reported in the past. The solution is compared to the arrangement that is presently used by the Coast Guard.