The aim of this research project was to produce a proof-of-concept system for the automated monitoring of target species of bird and bat in Australian cotton landscapes. The project consisted of three key components: 

1. Identify species of bird and bat that hep farmers manage on-farm biodiversity 
2. Create automated systems capable of accurately identifying the target bird and bat species 
3. Develop acoustic recorders for deployment in cotton growing regions that automate monitoring of target biodiversity. 
4. Create an online system that allows growers to visualise their in-farm biodiversity 

We have created a series of machine-learning (AI) based systems (using convolutional neural networks) capable of identifying 6 species of bird and 10 species or species groups of bat that inhabit the cotton growing regions of Australia. Each species has been selected because of the ecological and behavioural role it plays in the landscape. These recognition systems have been embedded in a field-deployable, rugged, automated, solar-powered, internet-connected audio monitoring device with results uploaded to the cloud. Once in the cloud, farmers can view biodiversity information relevant to their farm via their web browser. Farmers without internet access on their farms can use small, cheap audio recorders that require a level of human intervention before biodiversity results can be made available. 

To our knowledge, we have created the first autonomous acoustic species identification system for birds and bats with online reporting in the world.