Artificial Intelligence for Shipboard Asset Failure Prediction

Mitchell, Walter (SparkCognition, Inc.) | Rosner, Marla (SparkCognition, Inc.)


Walter Mitchell (V), Marla Rosner (V) SparkCognition Unexpected failures can have drastic consequences for ships, and taking a cue from other industries, operators are looking to implement predictive maintenance solutions using sensor data generated by shipboard machinery. One ship operator engaged with artificial intelligence (AI) company SparkCognition to pilot a predictive maintenance solution using historical data to predict failures of two critical assets: generator alternators and propulsion motors. SparkCognition's AI models were able to predict the failures months in advance. KEY WORDS: artificial intelligence; predictive maintenance; operations (shipping); planning; case study; machine learning INTRODUCTION This paper explores the use of artificial intelligence for predictive maintenance on maritime assets, specifically pointing to a case study predicting failures on generator alternatorss and propulsion motors for an operator of very high gross tonnage vessels. THE NEED FOR BETTER APPROACHES TO MAINTENANCE IN THE MARITIME INDUSTRY Any industrial machine--that is to say, any piece of equipment that rotates, creates pressure or temperature, or has flow--is subject to degradation over time.