What was it all about?

This project developed and tested an artificial intelligence (AI)-based monitoring system for counting banana spotting bug and fruit spotting bugs, two of the most damaging native insect pests affecting Australian horticultural crops.  

This work resulted in the development of Australia’s first specialised AI detection model for these pests, achieving a 93.8% detection score under controlled testing by AI training with thousands of images, and developed a field prototype capable of capturing images, monitoring sticky trap performance where banana spotting bugs were attracted to a sticky trap by a pheromone lure and transmitting data remotely from the farm.  

The project combined AI technology, computer vision and smart-farming hardware to create a pilot automated pest monitoring system, while a solar-powered AI spotting bug counting system was designed, built and deployed on a commercial avocado farm in North Queensland for a 10-week field trial.  

The project demonstrated the potential for AI-assisted pest monitoring in horticulture and established a foundation for future development of automated surveillance systems. Further field validation is required and there may be potential to help growers improve pest monitoring; however, the success of the AI counting method is also dependent on having effective lures to attract target pests to the sticky traps.