A facility location model for uncrewed surface vessels in the maritime survey industry

Abstract

The maritime industry is preparing for a future where human presence is no longer required on board of ships. This will revolutionize the global execution of maritime operations and consequentially introduce unprecedented challenges to the corresponding logistics. This paper presents a non-standard facility location problem (FLP) that arises in the maritime survey industry. The goal is to determine a number of uncapacitated facilities and assign a heterogeneous fleet of both remotely operated vessels and traditional vessels to the located facilities in order to serve the inspection demand of offshore infrastructures such as oil platforms and wind parks. The facilities serve as sites where vessels can refuel and accommodate crew changes. Both current traditional vessels and future uncrewed surface vessels (USVs) are considered in the heterogeneous fleet that respectively combines the distinctive routing behavior of centroid-like hubs and drone-like \textit{one-to-one} trips simultaneously. A mixed-integer linear programming (MILP) model is constructed to determine the optimal harbor locations and the associated vessel fleet size \& mix to perform the inspection demand of North-Sea assets over a multi-period time interval. The simulation results are reported for a real case study commissioned by geo-data company Fugro. The results of the model suggest that the effective establishment of facility locations and corresponding fleet allocation can reduce the total costs and environmental footprint of the survey operations in the North-Sea. This research provides a first piece of reflection regarding MIP problems for remote and autonomous operations in the maritime industry. The research pointed out that the complexity of USV operations in combination with traditional vessels is difficult to capture within an acceptably sparse facility location problem. Finally, this study identifies the stochasticity of inspection operations to be the most promising future contribution to automation research in the maritime industry.