NEWS – BatEchoMon: India’s First Automated Bat Monitoring System

HIGHLIGHTS

Overview

• Name: BatEchoMon (Bat Echolocation Monitoring)
• Purpose: Automated real-time detection and analysis of bat echolocation calls.
• Developed by:
  • Kadambari Deshpande (Postdoctoral Fellow, IIHS)
  • Vedant Barje (WildTech Project Lead, Wildlife Conservation Trust)
  • Under the guidance of Jagdish Krishnaswamy at IIHS, Bengaluru

Why BatEchoMon Was Needed

Challenges in Traditional Bat Monitoring

• Manual scanning of hours-long recordings.
• Time-consuming data analysis:
  • 30 GB of data generated per night (11 hours of recording).
  • Took 11 months to process data from just 20 nights.

Deshpande’s Experience

• Recorded bat calls in the Western Ghats for PhD research.
• Desired a system to automate data collection and analysis.

Technology Behind BatEchoMon

Core Components

• Ultrasonic Detector: Audiomoth (configured as a microphone).
• Processing Unit: Raspberry Pi microprocessor.
• Algorithm: Convolutional Neural Network (CNN) for species identification.
• Output:
  • Spectrograms of bat calls.
  • Audio recordings of identified calls.
  • Statistical analysis of bat activity by species and time.

Design Features

• Automatic Activation: Begins at sunset.
• Real-Time Processing: Detects, isolates, and analyses bat calls instantly.
• Enclosure Size: 200 mm × 80 mm × 80 mm.
• Power System:
  • Solar panel with battery backup.
  • Operates up to 8 days without sunlight.
• Data Transfer: WiFi-enabled.
• Modular Design: Customisable based on location needs.

Impact on Bat Research

Scientific and Ecological Significance

• Automates a previously manual process.
• Enables researchers to focus on ecological questions, not just data processing.
• Can be deployed across different regions to expand bat studies in India.

Development Journey

Inspiration and Collaboration

• Deshpande’s long-standing wish to build a localised monitoring tool.
• Collaboration sparked after meeting Barje.
• Iterated through multiple designs and components.

Cost and Accessibility

• Costs about a third of other advanced bat detectors.
• Designed to be affordable and scalable.

Challenges Ahead

Limitations in Bat Call Databases

• Currently identifies only 6–7 common Indian bat species.
• Need for robust, diverse reference datasets.

Future Goals

• Broaden species recognition across urban and forested landscapes.
• Extend beta testing to external users.
• Improve accuracy through national collaborations.

Looking Ahead

Improving Indian Bat Research

• Field still developing in India.
• Limited Indian submissions in global bat-call databases (ChiroVox, Xeno-Canto).
• Initiatives like the State of India’s Bats workshop are fostering collaboration.