At the recent TOC Americas conference in Panama, John Lund from Visy gave a presentation on artificial intelligence (AI) for gate and yard automation, where he discussed how Visy is able to use cameras to identify and track assets over their journey through a container terminal.
Knowing where an asset is at all times can eliminate the need for separate technology systems that identify assets when they arrive at handoff points where containers are exchanged from one piece of equipment to another.
While the level of technology varies, most terminals with process and equipment automation in the yard use OCR technology at the gate and, increasingly, on STS cranes, to identify containers coming into the facility. Inside the yard,
most container position information is generated from the location of container handling equipment, and by recording the handoff of containers between that equipment. In the case of road trucks, terminals have developed a variety of systems, including using an RFID tag or a driver ID card to identify a truck under a yard crane, and then associate that data with a container number.
At TOC Americas, Lund presented Visy Area, which takes a different approach by using OCR and other camera technology to identify containers at access points and continue to monitor their entire journey throughout the port. The
transaction process at container handoffs between equipment is automated with spreader OCR that reads actual container numbers.
For positive container identification at handoff points, Visy has developed TopView, a camera system for spreaders that takes images for OCR processing automatically when the twistlocks are activated. With TopView, every CHE
spreader becomes a data collection point for capturing both the container ID from the number on the roof, and an image of the roof itself for damage inspection purposes. This greatly simplifies the container identification process in the base of a truck with two 20ft containers, or a twin 20 yard move.
Lund explained how Visy has been able to combine AI with image capture technology to track and trace everything on the terminal using cameras. Using AI, Visy software can identify vehicles by counting the number and position of axles, the location of containers, the presence of a seal and IMO labels, container door orientation and other identifiers. These enable the system to detect, identify and follow a truck or any other piece of equipment as it moves around the terminal.
Compared to other automation systems, Visy notes that Visy Area utilises cost-effective, lightweight infrastructure that is highly accurate, and easily scalable. Furthermore, using multiple cameras plus OCR readers, Visy Area increases redundancy and the number of checkpoints that identify an asset on the terminal. Working as a system, Lund continued, the cranes and checkpoints on the terminal are constantly monitoring, updating and “self correcting” the data during operations, which can help reduce the number of exceptions caused by failure to identify a container or an asset.