Leveraging the Internet-of-Things for low-cost pest monitoring
Pest management is a crucial component of agriculture, but it is generally difficult and costly to obtain timely, reliable decision-making data due to the labour-intensive manual monitoring methods used. This makes it challenging to plan effective pest interventions, and also evaluate their success.
This project aims to address these issues through development of autonomous devices with the ability to detect and monitor pest animal populations in real-time. Based on acoustic rather than visual data, these devices will use deep-learning artificial intelligence methods to recognise acoustic signatures from pest animals, enabling detection and estimation of population densities. Use of acoustic data and deployment through Internet-of-Things sensor nodes mean the devices can monitor continuously, day or night, at a lower cost than visually-based methods. Initially, the devices will be trained to detect rodents, but have potential for application with a wide range of agricultural pests.
Professor Bernd Meyer