90% of the world's cargo is transported in maritime containers, but only 2% is physically inspected by customs authorities, opening the possibility for illicit activities.
Today, it is widely believed that the only viable way to control containerised cargo is through information-based risk analysis. In this way, it becomes possible to target high-risk shipments and proceed with physical checks, only where needed. With the research in the ConTraffic project, the JRC aims at creating novel techniques and methods for a more effective risk analysis in the domain. The research focuses on investigating the value and usage of raw data on the movement of containers, previously not systematically used by customs authorities. Through deep-web data mining, semantic data integration, sequence data mining, container itinerary analysis, semantic trajectory clustering and statistical analysis, the JRC's Contraffic system not only collects and creates a historical data warehouse of container itineraries, but also generates meaningful risk-related information for domain experts. The results of the project are made available to authorised EU customs officers and security authorities in Europe in order to demonstrate the value of the approach and improve it further.
The JRC, together with the European Antifraud Office (OLAF) and the Directorate General for Taxation and Customs Union (DG TAXUD) is running two pilot projects involving many Member States. One is concerned with systematic cross-checks of the origin of goods inside import declarations to detect customs fraud. The second one is dealing with systematic analysis of pre-arrival data in near-real-time in combination with Contraffic container trips data and aims at developing risk indicators for the safety and security of cargo entering Europe.
In line with the EU Strategy and Action Plan for customs risk management (with Annex) and based on its expertise in container movements, JRC also advises OLAF and TAXUD in proposing relevant legislation on reporting by the logistics industry.