Around 90% of the non-bulk cargo is transported in containers travelling by sea. In 2012 the estimated number of containers used for these operations was about 20 millions. Because of the huge number of the containers and the fact that inspection are expensive and create delays, only 2% of them are physically inspected by Customs Authorities. This fact is paving the ground to their potential usage in a number of illicit activities, including commercial frauds, illegal traffic, and even terrorist attacks. That’s why new methods for monitoring and analyzing the container movements are studied which could help the Authorities to target suspicious containers and inspect them. Container route based analysis is such a solution which has been considered in a series of studies [1-3] as a key factor in identifying potentially suspicious consignments. An essential part of container route analysis is the visualization technique which is widely recognized to be very powerful in anomaly detection. In addition, visualization methods take advantage of human abilities to perceive patterns and to interpret them, which can be critical in complex situations, especially when the dataset is multi-dimensional, massive and dynamic. When the dataset is really big, it is very difficult to assess what kind of data analysis techniques could be used for transforming the data. In such cases, the visualization of the data could help significantly by giving an initial general view of container’s behavior. In certain cases, this can also indicate abnormal movements, i.e. circles in container itineraries or unnecessary transshipments.
ConTraffic is a unique technology developed at JRC to screen data on global maritime container movements to detect potentially suspicious consignments. It gathers automatically container movements’ data from a number of on-line sources and features facilities to target suspicious containers based on analysis of their itineraries. ConTraffic provides several online services for data analysis and presentation of the data in textual shape. Recently the information visualization techniques are used successfully in combination with data analysis techniques to explore and visual analyses of large geospatial data. This is because a proper visualization of container traffic information stored in ConTraffic DB could become a powerful online service for better understanding complex container trajectories. Such a service could allow automatic pre-processing and presentation of the data in such a way that an expert user can identify complex irregularities and discontinuities. In this context, we propose a web-based prototype for geographical visualization of complex container traffic information that circumvents the problem of huge data size and dimensions by exploiting perception capabilities of human visual systems and needs.
The paper is organized as follows: Section 2 gives a short overview of the web-based visualization for data analysis, Section 3 gives a description of the ConTraffic System and DB, and Section 4 presents the initial requirements, proposed architecture of the prototype, details for the necessary data acquisition and manipulation and description of the realized user interface for data presentation. Finally, in the next two sections, we make a short discussion of the advantages, limitations, usage statistics of the implemented prototype, presenting future improvements and conclusions.