This work explores probabilistic deep learning models as offshore farm-wide virtual load sensors, including Bayesian neural networks, Monte Carlo dropout, and deep neural network ensembles. The aim is to develop models offering uncertainty-aware predictions of damage equivalent loads using SCADA and accelerometer data. This study uses the data from a Belgian offshore wind farm’s five fleet-leader turbines. After training the neural networks with one year of collected data, these models are deployed to another year, facing out-of-distribution data due to changes in operational conditions. The analysis assesses generalization and uncertainty quantification abilities, providing insights into their strengths and weaknesses. Ultimately, our work supports the industrial adoption of probabilistic deep learning virtual monitoring models, enabling informed asset management decisions with predictions and uncertainty measures.

The application of Deep Reinforcement Learning (DRL) for the management of engineering systems has shown very promising results in terms of optimality and scalability. The interpretability of these policies by decision-makers who are so far mostly familiar with traditional approaches is also needed for implementation. In this work, we address this topic by providing a comprehensive overview of POMDP- and DRL-based management policies, along with simulation-based implementation details, for facilitating their interpretation. By mapping a sufficient statistic, namely a belief state, to the current optimal action, POMDP-DRL strategies are able to automatically adapt in time considering long-term sought objectives and the prior history. Through simulated policy realizations, POMDP-DRL-based strategies identified for representative inspection and maintenance planning settings are thoroughly analyzed. The results reveal that if the decision-maker opts for an alternative, even suboptimal, action other than the one suggested by the DRL-based policy, the belief state will be accordingly updated and can still be used as input for the remainder of the planning horizon, without any requirements for model retraining.