Visy reports strong demand for new OCR use cases

Visy reports strong demand for new OCR use cases
Visy's cameras in action

Terminals have re-evaluated their business processes during the pandemic, resulting in significant demand for new use cases according to optical character recognition (OCR) and gate operating system (GOS) specialist Visy.

Visy’s Automatic Damage Detection System (ADDS) is the world’s first such system using vision technology.

Traditional OCR systems extract data such as license plates numbers, container codes, seal presence, door direction and hazardous good labels from camera images when a vehicle drives through a portal.

Using advanced vision technology developed in-house, the ADDS also extracts damage condition data from the same portal and the same images, identifying bulges, dents, extreme rust, and structural issues.

John Lund, global sales and marketing director, at Visy, said: “The use cases are only limited by your imagination. For example, when a damaged container enters the terminal, the system will treat it as an exception handling event and redirect the truck for further inspections or repairs.

“The system can also be designed with a web portal for container owners to log in and get information about the condition of their cargo. By automatically digitising the condition of cargo, new possibilities open up for terminals in terms of how quickly and efficiently they can move cargo and enhances services to their customers.”

The ADDS user interface

The technology provider is currently working on a solution called Visy AREA, which uses overview cameras, such as a CCTV system, to automatically detect, track, and trace all vehicles inside a terminal.

The location information on assets creates new automation possibilities for the operation while the camera network extends a traditional Visy GOS to provide data on identity and location of each vehicle as it travels through the facility.

Within the coverage of the overview cameras, the system provides an overlay of each vehicle location on top of a facility map and allows users to define virtual checkpoints (geofences) to the map.

Every vehicle passing through a virtual checkpoint generates an event like a regular event happening at a physical checkpoint (e.g., an in-or-out gate).

Moreover, Visy AREA features a dedicated Zone View, where a summary of vehicles within defined zones (e.g., a parking place) is shown alongside the map view.

Lund observed: “For operators, this means that with a CCTV camera infrastructure (that may already exist), they can automate processes based on what the cameras see and report on asset locations. AREA eliminates the need for physical checkpoints or roaming clerks with tablets or clipboards.

“The system sees everything that is happening in the facility and updates third party systems and operators in real time.”

Meanwhile, Visy’s spreader product TopView, which was launched in 2019, has quickly become a standard offering on all ship-to-shore (STS) crane OCR systems.

Lund explained: “The benefits are too big for operators to ignore. The spreader is the common denominator in container terminal operations. By turning the spreader into a smart device (i.e., data collection point) the operator receives digitised data every time a container is moved.”

For example, as soon as the twistlocks engage, TopView knows the box ID and will verify the move with the TOS.

In quayside operations, the system also collects damage inspection images of the roof of the box before the terminal even offloads it from the ship.

This can help terminals to refute damage claims because they can prove that the box was damaged before they hooked it.

Visy’s investments in R&D have helped to generate advancements in its dynamic neural network, which has enabled TopView as a standalone product to reach over 97% accuracy from raw reads.

Lund added: “The financial implications of the difference between 95% accuracy and 98% accuracy can be astonishing for a terminal. If you think about the terrible condition of container codes on the roof of a box, it’s amazing that the system can get such great results.

“Without these advanced vision technology tools, features such as automatic damage detection also would not be possible.”

Through the pandemic, the company has delivered on projects by utilising its local partners to perform physical work in situations where its Finnish team have been unable to travel to a customer site, instead remotely configuring the system.

Visy’s systems use commercial-off-the-shelf (COTS) hardware untied to specific brands, enabling it to source from different companies that still meet project requirements when a specific camera brand is out of stock.

Additionally, all its software including the OCR and GOS ecosystem is made in-house by Visy’s engineers and designed to be scaled.