Maritime is the vein of transportation in the archipelago area. Efficient maritime transportation poses a challenge to connectivity in regard to the archipelagic hurdles. Enhancing ship fuel efficiency is a valuable idea to address this challenge from an economic perspective and in the environment. This research aims to develop a comprehensive combination model from speed optimization and the Flettner rotor (FR) technology. Speed optimization starts with forecasting the speed using machine learning and optimizing the speed by reducing the anchoring time with the Just In Time (JIT) method. The FR technology focuses on harnessing the fluctuating wind in an archipelago area and changing it into power. The data needed for this calculation comes from the fusion of an Automatic Identification System (AIS) and weather data.

The research begins by designing the combination method of data fusion, speed optimization, and FR technology. In terms of the data fusion model, starts by identifying AIS and weather data as potential data sources. These data are selected based on their relation to fuel consumption. The data characteristics differ temporally. AIS data consists of point data every few seconds, while multiple weather data is in grid format, with the smallest time unit being hourly. Additionally, vessel movement data, such as heading, needs to be adjusted, making the data fusion process more challenging. The next step involves the design method of calculating speed optimization by combining speed prediction using machine learning with the JIT method to reduce anchoring time during ship operations near ports. The prediction approach with machine learning will always yield unique and specific techniques, heavily reliant on the patterns, quantity, and relationships among the available data. Subsequently, the design framework of a FR is installed on the ship to harness wind energy into propulsive power. The choice of an FR is based on its adaptability to various wind directions and compact dimensions, enabling easy installation on smaller ships. The combination method is tested using the experimental analysis using the real data and with the help of Python programming language.

This research concluded that the combination approach increased fuel efficiency significantly in the archipelago region. Furthermore, merging AIS and weather data produces a precious database for fuel efficiency calculations. Therefore, this method is highly adaptable and can be applied to different types of vessels and routes.