Marine fisheries are acknowledged to be some of the most important food resources for countries around the world. However, the issue of fishery sustainability has now become a key concern around the world. As reported by World Wild Fund for Nature (WWF), cod are currently at risk from overfishing in the UK, Canada and most other Atlantic countries. Global cod catch has suffered a 70% drop over the last 30 years, and if this trend continues, the world’s cod stocks will disappear in 15 years.
Illegal, unreported, and unregulated (IUU) fishing is one of the major threats to the sustainability of ocean fish resources. As estimated by National Oceanic and Atmospheric Administration (NOAA), IUU fishing produces between 11 and 26 million tons of seafood annually, representing as much as 40 percent of the total catch in some fisheries. The driver behind IUU fishing is high economic profit and low chance of seizure.
It is impossible to maintain a 24/7 presence to prevent IUU fishing everywhere due to the limited asset patrolling resources. Hence the allocation of the patrolling resources becomes a key challenge for security agencies like USCG. Research within this project aims to address the problem of deriving accurate patrol schedules for the US Coast Guard. To achieve this, our aim is to develop fast and effective techniques to solve security games defined over continuous spaces, where different types of defenders and attackers might co-exist and potentially coordinate their behavior.
Game Theoretic Fish Patrol Schedule Model Overview
The Game Theoretic Fish Patrol Schedule Model casts the interaction between the USCG (defender) and the illegal fishing boats, i.e., lanchas (adversaries), as a repeated Stackelberg game. Real-world data provided by the USCG is used to estimate the parameters of the game, including the behavorial model of the Lancha adversaries. The Stackelberg game can be solved using the MIDAS algorithm which computes the defender’s strategy, i.e., a randomized patrolling strategy. The defender strategy is then used to generate daily patrol schedules for the USCG assets. Executing these patrol schedules produces more Lancha data which allows the entire process to be repeated.