Smart EarTag for Livestock

JCU was awarded an Advance Queensland Innovation Partnerships (AQIP) grant to investigate a machine learning approach to terrestrial based geolocation with CSIRO and Ceres Tag. The project saw the development of an ultra-low power ear tag capable of delivering hourly location updates, and lasting years on a single battery charge with a range of at least 10 kilometres.

While time of arrival triatelation is a well understood approach, our system corrects for multi-pathing errors by applying machine learning neural networks using real time fixed location and environmental data.  We focused on freely available spectrum to reduce cost to end users, with location resolution of ~30 meters and a range in excess of 10 km.

This $1.5 million project developed new technologies for the Australian livestock industry. Our industry partner Ceres Tag is focused on providing a real world solution for the cattle industry that will change the way producers view their herd and stations. With the aid of CSIRO our researchers will develop new machine learning approaches to correct for errors in traditional terrestrial location systems.

Key researchers:

• Professor Ian Atkinson
• Professor Wei Xiang
• Dee-Ann Dunworth
• Dr Scott Mills
• Dr Bronson Philippa
• Nigel Bajema