Autonomous and Teleoperated Technologies for Comprehensive Rainforest Biodiversity Monitoring

摘要

This paper introduces a novel approach to rainforest biodiversity monitoring that integrates advanced remote sensing technologies with in situ data collection methods based on the team <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Welcome to the Jungle’s</i> submission to the XPRIZE Rainforest Competition. In this approach, an aerial drone collects multispectral, LiDAR, and thermal imaging data to accurately map canopy structures and biodiversity indicators while a second drone deploys and then later retrieves remote sensor units capable of collecting in-situ data throughout the many layers of the canopy. These sensor units are equipped with audio, camera, and environmental DNA (eDNA) traps. The system was field-tested in the Central Catchment Nature Reserve in Singapore, the OSA peninsula in Costa Rica, and along the Tumbira River, an offshoot of the Rio Negro River in Amazonas, Brazil. In the competition finals in Brazil, the system successfully collected data over 1km<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> of the Amazon rainforest within 24 h, followed by biodiversity analysis lasting 48 h. The paper focuses on the technology used in the data collection aspects of the competition and not the 48 h data analysis portion. Results from tests demonstrate the effectiveness of this system in overcoming traditional limitations of rainforest surveying, enabling unprecedented spatial and ecological data resolution. The research highlights the potential for enhanced ecological monitoring and conservation efforts through the combination of autonomous robotics and innovative sensor technology.