Machine Learning Based Approach to Predict Short-Term Fuel Consumption on Mobile Offshore Drilling Units

The application of machine learning models for optimization and improved decision-making has great potential in the drilling industry. This paper demonstrates a model for predicting fuel consumption on a Mobile Offshore Drilling Unit (MODU) with a Multi-layer Perceptron (MLP) artificial neural network. The model is proposed as a tool for setting fuel consumption related performance goals for offshore personnel on a MODU. Operational and environmental data have been used as input variables for the model, with a dataset split into 80% training set and 20% test set. The highest performance is obtained with three hidden layers with 38 nodes each. The Adam solver performs better than the Stochastic Gradient Descent (SGD) solver for weight optimization, and the best a parameter for the L2 regularization term is 0.0001 with the Adam solver. The MLP regression model predicts fuel consumption for the test set with a Root Mean Squared Error (RMSE) of 0.0770. This result indicates that artificial neural networks and the MLP regressor is a suitable algorithm for predictive modelling of fuel consumption on a MODU.