Trellis Data, in collaboration with the Australian Department of Agriculture, Fisheries and Forestry, has successfully completed a 5-month pilot trial of their AI-powered Biosecurity Automated Threat Detection System (BATDS) at the port of Brisbane. The trial aimed to protect Australia’s agriculture sector, which is worth $90 billion, from pests that are introduced through the country’s ports.
Australia faces constant threats from pests that enter the country through shipping containers. To address this challenge, the department commissioned Trellis Data to develop an AI-based surveillance system capable of scanning all incoming containers. Trellis Data used its Trellis Intelligence Platform to create bespoke Object Detection Models and integrated them with advanced camera management technology to create the Biosecurity Automated Threat Detection System.
During the trial, the AI model evolved and was able to identify approximately 58% of container anomalies that were identified through manual inspection. Notably, it successfully detected 63% of soil detections during the final reporting period. This technology has the potential to not only safeguard the agriculture sector but also protect goods, identify illegal items, and enhance overall border protection.
The BATDS system scanned over 48,000 containers for biosecurity risk materials using 35 strategically placed cameras on five cranes at the DP World facility. The system seamlessly integrated with existing processes, allowing for swift container scanning without disrupting the flow of goods. As part of the pilot, 1,300 high-risk containers were manually inspected by the department, and the results were compared with the detections made by the BATDS system.
A continuous improvement process was developed with entomologists from the department to fine-tune the AI model, resulting in a highly effective machine learning model for biosecurity screening. The model was able to filter out non-threatening objects such as rust, grease, and container damage, leading to more accurate detections.
In order to track all detections comprehensively, a separate model was trained to read container IDs on challenging surfaces. The Trellis Data system design allowed for wireless streaming of all captured images from the cameras to servers, enabling the highly mobile cranes to move seamlessly throughout the port.
Tim McLaren, the Head of Communication at Trellis Data, believes that this pilot project is a success and demonstrates the immense potential of AI-driven systems in enhancing biosecurity screening at scale. McLaren stated that they are now looking for funding to implement this technology on a larger scale and are interested in partnering with other countries such as the US to further test and improve the system.
With ongoing improvements, Trellis Data aims to significantly enhance biosecurity screening at all port facilities in the future. This advancement will provide greater confidence in the biosecurity status of containers entering Australia, further safeguarding the country’s vital agriculture sector.
Link to BATDS overview video: https://www.youtube.com/watch?v=QxJx0-3589c&t=1s