Marine Vessel Telemetry Data Processing Using Machine Learning

Herry Susanto, Gunawan Wibisono


In Indonesia, one of the causes of the high cost of fuel in the shipping industry is theft and misuse of fuel. This happened because ship management center unable to monitor all the activities of the ship when the ship sailing in the middle of the ocean. Lately, ship monitoring through the latest technology are being carried out, one of which is the Machine to Machine (M2M) based Vessel Monitoring System (VMS) technology. The development of VMS and telemetry technology has enabled monitoring of engines and fuel consumption of ships in real time. The problem with this VMS system is that there is still a dependency on the analysis of experts who need a long time to analyze various parameters of existing telemetry data, which lead to inaccuracy and delay in anomaly detection. This study conducted a statistical analysis of telemetry data, especially in ship movement and machine activities, and then designed the fuel consumption regularity classification system with the Naive Bayes and Logistics Regression. Naive Bayes method was chosen because it can produce maximum accuracy with little training data, and Logistics Regression was chosen for its simplicity and excellent results in prediction of numerical and discrete data. The results of this study indicate that telemetry data from the VMS system can be used to detect irregularities in Fuel consumption. Tested with selected data, Naive Bayes classification accuracy in irregularities detection is up to 92% while logistic regression is up to 96%.


Vessel Monitoring System; Machine Learning; Naive Bayes; Regresi logistik; GPS; M2M

Full Text: PDF


  • There are currently no refbacks.