Development of a novel Decision-Making tool for vessel efficiency optimization using IoT and DL

Authors:
Publisher website
Google Scholar
Abstract
The era of the Internet of Things and big data has completely changed the landscape of many scientific disciplines. Embracing the big data revolution, machine learning techniques seem to have substituted traditional physics-based models due to their effectiveness and flexibility. According to Rolls-Royce [1], the smart ship will revolutionize the landscape of ship design and operations, the same way smartphones did our daily life. These enhanced data-driven analytics yield substantial benefits in terms of cost savings by limiting the degree to which human factor intervenes to perform certain operations, and also by providing reliable and accurate information in real-time information regarding the operational health of the machinery on board, ensuring the safety of the crew as well as the reliability of the equipment. [2]. Despite its importance, the majority of efforts towards this direction are limited to academic efforts or black-box approaches. In the current work, a distributed software platform that augments the decision-making process within the shipping industry is critically presented. Key technological pillars are Big Data technologies, data fusion, and large-scale deep learning-based processing techniques. Evaluation of the platform, based on actual case data, provides evidence of its capacity for diagnosis and anomaly detection on the operation of critical subsystems.