Port Terminals: Tracking containers within the terminal yard.
Optical Character Recognition
Port congestion results in a chronic gridlock at ports across the globe,. posing an economic threat and barrier to international trade. This emanates from the port operators’ inability to locate missing containers within the port yard terminal. Ships are stranded offshore for days or even weeks, waiting to berth. Terminal yards are clogged and trucks are waiting in line for hours to pick up a single container, causing shipping delays for customers throughout the entire supply chain. This becomes a significant trade barrier for both exports and imports with a corresponding negative impact on the economy. According to the National Cargo Security Council (NCSC), the maritime industry loses $50bn annually due to cargo theft and cargo loss.
Sub-Saharan Africa’s largest and busiest port terminal, Durban Container Terminal (DCT) was the testing ground for a container tracking application which utilized Optical Character Recognition (OCR) as the leading technology. The objective of this case study was to ascertain the feasibility of tracking containers, using container serial numbers, within the container terminal. This was largely driven by our finding from our preliminary design and selection of a solution for cargo theft and loss. It was clear to us that attaching a tracking device to a container would not be feasible for port terminals due to i.) Regulatory restrictions ii.) Technology restrictions iii.) Port terminal operational restrictions and that the most optimal solution would be one that required no physical attachment to the container itself.
Underpinned by OCR technology, the selected solution consisted of a proprietary wireless edge camera, strategically placed on container handling equipment (i.e. straggle carrier, quay crane etc.) to capture the container serial number and georeference it. This data would be visualized on a cloud based web-application, allowing key stakeholders the ability to search a container's location from anywhere in the world.
In the three months test period, we observed an average accuracy of 85% in container serial capturing and character recognition. Paired with a proprietary latitude, longitude and altitude position logging technique, with a 50 cm accuracy. We are confident that this tracker can be used without any physical attachment to the container.