Time-series analysis approach for improving energy efficiency of fixed-route passenger vessel in short-sea shipping

• Develop a data-driven model to improve voyage energy efficiency in short-sea shipping. • Propose an efficiency score combining fuel consumption and voyage duration metrics. • Implement clustering to classify voyages by efficiency, aiding optimization and insights. • Evaluate four time-series models for speed optimization using real-world voyage data. • Demonstrate practical applicability for fixed-route vessels with limited optimization options. Improving ship energy efficiency is critical for reducing operational costs and meeting increasingly stringent climate regulations. However, existing approaches often target variable-route vessels and overlook the specific operational constraints of short-sea, fixed-route passenger services. This paper presents a novel data-driven framework that applies time-series analysis techniques to optimize the speed profiles of fixed-route passenger vessels. The framework introduces a spatiotemporal aggregation method for fusing operational, environmental, and navigational data from onboard IoT devices and external sources, enabling the derivation of a new efficiency score that jointly considers fuel consumption and voyage duration. By applying clustering techniques to this efficiency metric, voyages are categorized for targeted optimization. The framework evaluates four distinct time-series models across a dataset of 1755 real-world voyages collected over 15 months in southern Sweden. The findings highlight the effectiveness of time-series analysis approach for optimizing vessel voyages within the context of constrained landscapes, as often seen in short-sea shipping.