The Amazon rainforest constitutes the world’s most intricate ecosystem, harbouring countless species of flora, fauna, and wildlife, many of which remain unexplored by scientific inquiry. Between 2022 and 2023, the Amazonian region witnessed a staggering rate of deforestation, with an average daily loss of approximately 10,000 acres, according to the Monitoring of the Andean Amazon Project.
Approximately 10% of the Amazon rainforest lies within Colombia’s territorial boundaries, where innovative collaborations are leveraging cutting-edge cloud and artificial intelligence technologies to develop a pioneering approach to curbing deforestation, mitigating the effects of climate change, and safeguarding the region’s rich biodiversity, as well as Indigenous communities, from harm.
According to Luz Marina Mantilla, director of Colombia’s Instituto SINCHI, preserving the Amazon region’s integrity is crucial because once it’s compromised, the ecosystem’s protection, biodiversity, and lush vegetation are all at risk of disappearing. “That’s something we’d rather avoid happening.”
Scientists at the prestigious Lab join forces to pioneer a groundbreaking endeavour that leverages cutting-edge AI innovations to monitor deforestation and safeguard the precious biodiversity of ecosystems, seamlessly integrating satellite imaging, camera traps, and bioacoustic monitoring technologies.
According to Pablo Arbeláez, director at the Centre for Innovation and International Analysis (Cifonía), expertise has the potential to significantly contribute to preserving the Amazon.
Through the integration of data, machine learning, cloud technology, and data science, researchers can identify patterns of deforestation more quickly, enabling swift action to be taken in vulnerable regions before irreparable harm is done. The Audio Language Model, known as CLAP, is a custom-built multimodal language model developed by Microsoft. The picture fashions are trained from open-source architectures and employ a framework developed by Microsoft, specifically PyTorch Wildlife.
According to Juan Lavista Ferres, Microsoft’s AI for Good Lab Company Vice President and Chief Information Scientist, “The initiative effectively demonstrates the synergy of collaborative efforts from personal sector entities, NGOs, universities, and governments.” This instance illustrates how AI expedites and supports the vital efforts of conservationists. We’ve achieved significant momentum through our joint effort, and I’m excited about the prospect of further impact.
Over the past year, significant updates have been implemented in Undertaking Guacamaya, significantly enhancing researchers’ capabilities and opportunities to track and analyze deforestation patterns.
- Prior to this replacement, imagery was provided on a monthly basis. With real-time updates provided daily, scientists can now trigger timely alerts utilizing high-resolution imagery from satellites constantly monitoring the Amazon.
- Researchers studying newly discovered Amazonian species now leverage a bespoke AI model tailored to the unique characteristics of the rainforest, enhancing accuracy and enabling them to identify species up to 10 times faster.
- IDEAM, Colombia’s official provider of deforestation research, has partnered with Undertaking Guacamaya to incorporate innovative visual storytelling into its reports.
- The venture has launched an open-source platform specifically designed for creating, modifying, and sharing highly effective AI conservation models.
“With this confluence of data, institutions, and expertise, we urge the nation to take bold action in making crucial decisions on how to conserve and protect ecosystems,” asserts Hernando García Martínez, general director of the Instituto Humboldt. “It’s essential that people understand the true value and significance of nature.”
