In Japan, it is becoming difficult to move freight long distances using only trucks, owing to a shortage in truck drivers. In this context, there is a tendency to also actively utilize ferries or roll-on/roll-off (RORO) ships. Intermodal freight transport uses two or more suitable modes of transport from the origin of goods to the destination. In order to promote intermodal freight transport, it is necessary to analyze factors such as the characteristics of transported goods, transportation services and fares, and lot sizes. In this study, we focused on ferries and RORO ships in Japan, and attempted to identify cargoes transported using different modes of transport, based on the net freight flow census. In conclusion, high volumes of “agricultural and fishery products,” “metal machinery,” “chemical industrial,” “light industrial,” and “miscellaneous manufacturing products” can be transported using ferries or RORO ships. Specifically, there is a tendency that products are segregated by transportation mode such as a truck, or a ferry and a RORO ship.

Automatic identification system (AIS) data are used to analyze vessels’ positions or maritime traffic. Recently, satellites are being adopted for gathering AIS data (satellite AIS). A satellite AIS can gather AIS data from all over the world, including the center of the ocean. However, because a satellite moves around the world every day, there is a very short window of time that the satellite can receive signals from some specific area. Furthermore, because a satellite AIS receives signals simultaneously from a wider area than a terrestrial AIS does, the satellite AIS has lower rate of signal reception, especially near vessel congested areas. This may cause many unrecognized vessels. For this situation, this paper proposes a new method to estimate the number of unrecognized vessels based on a binomial distribution model. With this method, we are able to estimate the number of unrecognized vessels just from satellite AIS data themselves.