B.A.T. - Bio Acoustics Tracker
 The bird I identified in my garden during the project work, few days back. |  Black hooded Oriole detected from their calls in my garden! |
|---|---|
 Spectrogram of loud calls of a bird overlapped with some musical sound played somewhere externally. |  Cleaned Spectrogram of loud calls of a bird overlapped with some musical sound played somewhere externally. |
 Continuous bird chirps spectrogram |  Continuous bird chirps spectrogram cleaned |
 Cleaned Spectrogram of a bird calling in specific pattern |  Cleaned spectrogram of the Asian Koel bird calling |
The project Bio Acoustics Tracker, also known as, B.A.T. is an initiative to save the faunal biodiversity in urban areas by using technology. At the advent of urbanization and development of smart cities, we are abruptly destroying the urban ecology without knowing how it is hampering the environment and our entire ecosystem. To deal with urban development and conservation of biodiversity parallelly, we are developing a network of B.A.T. devices to perform acoustic monitoring.
Applications: Monitoring the variation of City biodiversity over time; Identification of endangered species in urban areas; Observing the city soundscape; Analysis of urban noise profile in smart cities.
Timeline: October, 2020 - March, 2021
Collaborator(s): Achal Nilhani
Theory behind Working: These devices will be implemented as a dense network throughout the cities and will continuously capture the acoustic data of that region, which will then be sent to a cloud server for processing. The power of Artificial Intelligence further helps in extracting the calls of different species (birds, insects, animals) from these long audio recordings. Based on this species data, project B.A.T. will generate a biodiversity map of different part of the city: this will help the authority identify locations with high biodiversity and accordingly plan to protect such areas even at the time of urban development. On the other hand, the continuous acoustic data will also produce a way of monitoring the variation in urban noise.
Data: The bird calls were identified from a 7.5 hrs long recording (from 12 midnight to 7:30 A.M.) with the recording system placed in my garden yard. Some birds can be identified quickly whereas some are yet not known to me. However, the algorithm we are developing will detect them all (and other species sounds also) to map the species in this area.