28th November 2023

We are delighted to receive the AWS UK Imagine Grant for our acoustic survey tool, BCT’s Sound Classification System (SCS). The grant will be crucial for the development of a new version of our SCS, hosted in AWS (Amazon Web Services).

When fully developed, the BCT’s SCS will fully automate audio-identification. It will be an end-to-end system - users can book equipment, upload files, classify bat-calls with AI, and create usable results all in one place.

Upscaling bat surveys

Critically, BCT’s SCS answers an urgent need to upscale bat surveys. Conventional bat surveys are time intensive. They require laborious field surveys and expert sound identification skills. Now, work made possible by the grant will allow us to rapidly scale up the data volumes flowing through the system. Using AI and cloud technologies we can achieve unprecedented efficiency and accuracy.

As a result, we will have bigger, better data and so better understand how bat populations and the ecosystems they rely on are faring. This means we can create a robust plan for the conservation actions needed in the face of global change. And because our data also goes to JNCC and Defra, our impact on UK policymaking is strengthened as well.

BCT win AWS UK Imagine Grant

BatDetect2 - the heart of the BCT's SCS - in action classifying Serotine and common pipistrelle bat calls. BatDetect2 was developed in collaboration with Kate Jones at UCL and Oisin Mac Aodha at Edinburgh University.

Removing barriers to surveying

Users will get bespoke reports about bat species in their area. Some can also get specialised training to verify identifications. Once they have that skillset, their work fine-tunes the algorithm and expands an open-source library of bat sounds (EchoHub) to create gold standard datasets.

BCT’s SCS was also designed to bring underrepresented demographics into bat surveying. Unlike traditional surveys, SCS users don’t need to watch bats in real time. So SCS removes barriers for those who can’t take part in traditional, physical surveys.

Surveys can also be done in urban areas, with sensors placed on balconies, windows or green spaces. Identification is also ‘gamified’ to appeal to younger demographics.

Meet the BCT scientist whose team develops BCT’s SCS

Find out more about audio-identification and the new technology driving it in our interview with Dr Lia Gilmour. Read: AI and audio-identification of bats: Lia Gilmour interview.