United States Department of Transportation
i
Use of Vessel Automatic Information System Data to Improve Multimodal Transportation in and Around Ports
-
2018-09-01
[PDF-2.60 MB]
Details:
-
Creators:
-
Corporate Creators:
-
Corporate Contributors:
-
Subject/TRT Terms:
-
Publication/ Report Number:
-
Resource Type:
-
Geographical Coverage:
-
Edition:Final
-
Corporate Publisher:
-
Abstract:Multimodal transportation is an evolving system in supply chain management and an effective approach for facilitating the movement of cargo when different modes of transportation are available and involved. Since one mode of transportation is usually insufficient for door-to-door transportation of cargo, multimodal transportation has become an important concept. Hence, it becomes necessary to transfer goods from one mode of transportation to another. For an effective multimodal system, the different modes of transportation involved require close coordination and precise synchronization, especially in terms of arrival times and cargo allocation (synchro-modality). The accuracy in arrival time of the vessel is most vital to imports, since it initiates the process of a multimodal transfer. Lack of certainty in estimated time of arrival (ETA) creates problems like delays and congestion at ports. It also leads to inadequate planning and resource management for port facilities and receiving modes of transportation. Vessel Automatic Information System (AIS) data provide vessels’ voyage information, including the ETA as determined by the vessel’s captain/operator. This information (the captain’s ETA) is manually inputted into the system and thus is subject to errors. Furthermore, captains sometimes forget to update such information, which affects the results of the analysis. Hence, the authors propose a way to generate ETAs from a system that does not require the captain’s ETA as input. This research describes an approach that generates the ETA of vessels to the port terminals by using machine learning and AIS data. The results of the analysis show that near-exact predictions can be achieved without prior estimations by vessel captains. The results indicate that the farther from the destination, the more errors are made in prediction. This is also evident in the comparison of prediction errors between Bayport and Barbours Cut, two container terminals in the Port of Houston. The analysis shows that predictions made at the terminal level are more accurate than at the buoy level. The ETA predicted from this approach provides an adequate timeframe within which terminal and trucking companies can plan for the vessel’s arrival.
-
Format:
-
Funding:
-
Collection(s):
-
Main Document Checksum:
-
Download URL:
-
File Type:
