EchoBank
In this Episode I visit Dr Kate Jones at the Zoological Society of London to find out about a bat call reference library they're developing called 'Echobank' which is being used in conjunction with the iBats monitoring program. Scientists at the Zoological Society of London are using Echobank to teach a neural network to identify specific bat species from the acoustic properties of their calls.
Although this may sound pretty niche, it's really cool stuff because the technology has much wider applications. Firstly the team are using it in combination with a smart phone app which can be used by anyone to take bat call recordings. In doing this the team hope to collect global distribution data for bat populations, which Dr Jones is hoping to use to determine whether bats can be used as a 'heart monitor' for the state of the environment. As bats represent one fifth of all mammalian species and exist in a huge range of habitats (from your local park to the tropics), changes in their global distribution could be used to monitor the impact of climate change on the natural world.
Secondly, the team are doing a lot of really important science engagement, working with ‘citizen science’ networks across Europe to gather data on bat populations. The team have been helping groups across Europe to develop and carry out their own monitoring programs, which feed data back to the iBats program. What’s great about this work is that it’s not only efficiently collecting distribution data, but also getting people interested in the state of their local environment. Standardisation is obviously an important factor when collecting data from multiple sources and with this in mind the team have developed the smartphone apps which can be used easily to take recordings and GPS data.
Finally (and what really excited me) was the suggestion that this ‘digital infrastructure’ could one day be adapted to identify any sound producing species from a recording taken on a smartphone. So imagine going out to your local park, taking a recording of a bird on your iPhone and getting probability results back on the identification of the species!
You can listen to the episode here:
[soundcloud width="100%" height="81" params="" url="http://api.soundcloud.com/tracks/11704888"]
Or listen to the edited version as part of Short Science episode 89
Special thanks to Katie Draper who helped out with this episode!