Replication Data for: Evaluating the effect of risk metrics for supporting operational decision-making by autonomous surface vehicles

This dataset contains replication data for the study reported in the paper “Evaluating the effect of risk metrics for supporting operational decision-making by autonomous surface vehicles” (submitted). The goal was to investigate how the choice of risk metric affects the behavior of an autonomous surface vehicle (ASV) that uses risk information from a risk model to support its decision-making during operation. This dataset contains data from path planning for an ASV in simulations and field trials. A case study of an ASV operating in cooperation with an autonomous underwater vehicle (AUV) was used. The ASV had to plan a path that considered the risk of collision with another vessel, grounding, and losing communication connection with the AUV. The risk estimate from a risk model was expressed using different risk metrics. The path planning was done using a method for dynamic probabilistic risk assessment named K-shortest paths probabilistic risk assessment (KPRA). The ASV was given the same task, i.e., planning a path between two points while avoiding undesired consequences, but with different risk metrics used to support the decision-making. The resulting paths chosen under the information from different risk metrics were recorded. Several datasets are included here, both from simulation studies, where the ASV, AUV, and a target ship (TS) were simulated, and from field trials with a real ASV. The datasets include the location, heading and speed of the vehicles, information about the plans selected by the ASV, and about the waypoints that were explored when searching for the plan. The area of operation was close to the Munkholmen Island in the Trondheimsfjord, Norway.