Harnessing nature sounds

A Low-Cost, Standalone Acoustic Monitoring System for Biodiversity Conservation 

Global biodiversity change poses a significant threat to numerous species, with many facing the risk of extinction, while also endangering the provision of crucial ecosystem services. Simultaneously, invasive species are spreading across the globe, contributing to biodiversity depletion. These changes, exacerbated by climate warming, entail substantial economic costs. To effectively manage and respond to these challenges and stay within safe operational planet boundaries, there is an urgent need for tools to quantify and monitor biodiversity. Widespread and continuous monitoring is essential but demands substantial logistical commitments. Current automatic acoustic monitoring systems offer vital insights into biodiversity patterns and changes. However, there is a gap in the ability to deploy, record, recognize, analyze, and report biodiversity data cost-effectively and continuously. This project aims to develop a low-cost, energy-independent hardware and software solution to measure and monitor acoustic information, delivering real-time insights into biodiversity patterns and changes, including species occurrence and variations in abundance.

The current state of biodiversity calls for innovative monitoring solutions that are cost-effective, continuous, and capable of providing real-time data. Existing methods, including manual observations and remote sensing, are often limited in their ability to capture biodiversity dynamics comprehensively. In recent years, automatic acoustic monitoring has emerged as a promising tool for monitoring biodiversity, as it can detect and identify species through their sound patterns. However, there is a need for a low-cost and energy-independent system that seamlessly integrates hardware and software to provide real-time information on biodiversity patterns and changes. Continuous monitoring of biodiversity is vital to assess the impact of these threats and guide conservation efforts effectively. Timely data on species occurrences and abundance variations are essential for evidence-based decision-making. As such, the primary objectives of this project are as follows:

i) To design and develop a low-cost system for acoustic biodiversity monitoring.

ii) To create a system capable of continuous data recording, species recognition, data analysis, and real-time reporting.

iii) To ensure the system is energy self-sufficient, reducing the need for frequent maintenance and power supply.

iv) To monitor and report on species occurrences, including non-native species, and detect changes in abundance patterns.

Most of laboratory work will be developed in FCUL facilities. Fieldwork will be performed in the Herdade da Ribeira Abaixo, the field station of FCUL, where the BASS project is hosted.

Significance of the Project

This project addresses critical issues in biodiversity conservation and management by developing an innovative solution for acoustic monitoring. The significance of this project lies in its potential to:

1. Provide real-time data on biodiversity patterns, helping researchers, conservationists and decision-makers to respond rapidly to changes.

2. Offer a cost-effective tool for monitoring biodiversity, making it accessible to a wider range of stakeholders (e.g., farmers, conservationists, researchers and/or government agencies).

3. Contribute to the management and mitigation of invasive species, aiding in the preservation of native ecosystems.

System Design

The design of the system involves selecting suitable hardware components and developing software for data collection and analysis. The system should be compact, durable, and capable of autonomous operation. Key components of the hardware include acoustic sensors (microphones) with high sensitivity and low power consumption; energy harvesting technology, such as solar modules, for energy self-sufficiency; data storage units with sufficient capacity to store continuous recordings; ability to send data remotely and automatically (e.g., using GSM technology); rugged and weather-resistant casing to protect the system in outdoor environments. On the other hand, the software should be able to provide real-time data collection; machine learning algorithms for species recognition; data analysis tools to detect changes in biodiversity patterns (performed in cloud-based tools); a user-friendly interface for system configuration and data access; and integration with cloud-based platforms for remote monitoring and reporting.

Key challenges to accomplish this project include the development (and/or adaptation/integration) of software to recognize and identify acoustic patterns, from insects and birds to human-made sounds. Another challenge is to ensure energy self-sufficiency. The system should incorporate renewable energy sources (solar panels), to ensure continuous operation without reliance on external power sources. Battery backup systems may also be included for uninterrupted monitoring during adverse weather conditions. Finally, data storage and transmission is a third major challenge. Recorded data must be stored locally and transmitted in real-time to a centralized database for analysis and reporting. Data compression techniques may be applied to reduce transmission costs.

Prototype Development

In the course of this project, a prototype of the monitoring system will be developed based on the design specifications. It will undergo rigorous testing in controlled environments to ensure functionality and durability. Then, the prototype will be deployed in various natural environments to assess its performance in real-world conditions. Data collected during field testing will be used to refine the system's algorithms and optimize its performance. Once the prototype is functioning, continuous data collection will be carried out to monitor species occurrences and abundance variations. Machine learning models will be trained to recognize species based on their vocalizations, and data analysis algorithms will be developed to detect changes in biodiversity patterns. Based on field test results and ongoing data analysis, the system will be refined and optimized. User feedback and stakeholder input will be incorporated to enhance usability and functionality. The monitoring system's performance will be evaluated based on criteria such as species recognition accuracy, data storage capacity, energy efficiency, and robustness in different environmental conditions.

The system will provide real-time data on species occurrences and abundance variations through a user-friendly interface. This information will be accessible to researchers, conservationists, and government agencies or policymakers. The system will track the presence and absence of target species, including non-native and invasive species, enabling rapid response to emerging threats. Changes in species abundance will be detected and reported, facilitating early intervention to protect biodiversity.

The Team

The project assembles a multidisciplinary team, each member bringing a unique set of skills and expertise crucial to its success. Fernando Ascensão, an ecology monitoring and modeling expert, will provide insights into biodiversity dynamics and ecological patterns. José Cecílio, a specialist in wireless sensor networks and the internet of things, leads the hardware development/integration, ensuring seamless data collection and transmission. Sara Silva has extensive experience in machine learning methods, including their application in remote sensing. Miguel Brito and Ivo Costa are photovoltaics technology specialists, who will focus on energy self-sufficiency, incorporating renewable energy sources into the system. This team embodies the synergy required to tackle the complex challenges of biodiversity monitoring and conservation effectively. 

Here are my team members