LED BY
INNOVATION PARTNERS
About
Tech4Nature Mexico works to accelerate the effective conservation and regeneration of biodiversity and ecosystem health, strengthening monitoring, conservation, and understanding of the impacts of climate change on ecosystems and priority species in the mangrove area of the Yucatán Peninsula.
Our mission at Tech4Nature México centers on understanding, preserving and restoring the Dzilam State Reserve to provide a secure sanctuary for a rich variety of plant and animal species.
Nestled in the northeastern region of Yucatan, the Dzilam State Reserve is a natural protected area with over 69,000 hectares that belongs to the municipalities of Dzilam de Bravo and San Felipe.
This reserve holds a special status as a critical wetland conservation site, boasting nearly 290 species of fauna intricately linked with over 300 flora species. It spans five distinct vegetation types, including coastal dunes, mangroves, petenes, along with vibrant aquatic flora in coastal lagoons.
The Reserve
The impact
2022-2024
.jpeg)
Automated species detection and identification
01
24 camera traps.
02
147 species identified in the reserve thanks to the project (40 of which are in some risk category).
03
Over 100,000 images and videos, over 600,000 audio files collected.
04
Automated analysis: from 3 years of work to just a few months.
Development of 2 pioneering identification algorithms
01
Development of 2 algorithms for identifying individual jaguars with over 90% recognition accuracy at the individual level.
02
9 jaguar individuals identified.
Data visualization and analysis platform
01
Development of the first platform to identify jaguars and support collective action and decision-making.
02
Development of reports with lessons learned, recommendations, and best practices for the ethical use of AI for biodiversity, promoting replicability and scalability.
03
Training of more than a dozen data engineering and embedded systems students to develop autonomous systems for biodiversity.
The impact of Tech4Nature Mexico goes beyond tangible benefits for conservation, community development, and the strengthening of public policies, positioning the project as a model to follow at the local, regional, and global levels.
Creation of Two Dashboards
An audio analysis platform developed by Rainforest Connection (RFCx) to identify, classify, and create biodiversity patterns in Dzilam.
Development of convolutional neural networks to label images based on the (non)presence of a jaguar
Creation of Two Algorithms
(pattern matching)
Acoustic Neural Network
95
Automatically validated species
60
Processed audios
Neural Network for Image Recognition
93%
Accuracy of the jaguar detection algorithm in images
Phase 2 Goals
2024 - 2026
Strengthen governance and replicability to increase conservation impact
Enhance the collaborative experience of the AI-based image recognition platform, with special attention to wild felines.
Expand local, regional, and international strategic partnerships to strengthen the initiative.
Expand support for the integration of reserves into the IUCN Green List.
Institutionalize multisectoral governance
The Biodiversity
Watch, listen, and see the biodiversity of the Dzilam de Bravo reserve!
PROGRAMME EXPLAINABILITY, TRANSPARENCY, INCLUSIVENESS, AND OTHER PRINCIPLES
AI Ethics is at the core of this project. Based on UNESCO's Recommendation on AI Ethics (in which C Minds' founder had the role of leading the Environmental chapter), geographic locations and other sensitive information is protected.
If you are a scientist and need access for scientific research and conservation purposes only, please email the project's data governance committee at regina@cminds.co
SUPPORTED BY
Our blog
