An integer programming approach to fisheries observer deployment

Fisheries observers - independent workers assigned to commercial fishing vessels – provide critical data on fishing activity and the state of fisheries. However, the logistics of deployment of observers can be challenging, often involving vessels docking at far-flung ports, high transportation costs, and potential for compromised observer impartiality. We develop an optimization-based approach for efficiently assigning observers to vessels, while meeting complex logistical and regulatory constraints. We test our model on commercial fishing schedule data, and demonstrate that this approach can reduce costs of observer transportation and risks of compromised observer impartiality, and quantitatively evaluate tradeoffs in large-scale deployment scenarios.