AI for Earth partners
Learn about our partners working on the front lines of sustainability.
eMammal uses Microsoft Azure to store and organise images from citizen-run camera traps while learning more about the world’s land mammals.Learn about eMammal
What does it mean for the ecosystem when invasive feral pigs tear through an area? What is the correlation between deer populations and spreading disease? Researchers hope to answer questions like these by studying local wildlife, but their organisations are critically underfunded – and they need help. Citizen scientists have the means to record valuable photos and data to help scientists tackle these complex conservation challenges. Their contributions are crucial to understanding and rehabilitating the world’s land mammal population.
eMammal is a data management system and archive for camera trap research projects. The platform makes wildlife photography accessible, opening the door for citizen scientists to contribute their own images to environmental research efforts. Contributors upload images captured from at-home camera traps to eMammal, where the images are tagged and organised by species type. eMammal uses Microsoft Azure Machine Learning to categorise its massive library and guide contributors to record more accurate results. Wildlife researchers and organisations can then access these images to gain a better understanding of wildlife populations and answer critical questions about animal behaviour, reproduction, ecology, genetics, migration and conservation sustainability.
A window into the secret life of ecosystems
Camera traps allow scientists to observe wildlife in their native environment. Photo data collected from thousands of cameras can allow scientists to understand complicated patterns of migration, changes in populations, behaviour in the wild and animal interaction. AI helps process this vast amount of data, allowing scientists to better understand these complex interactions and conserve ecosystems at risk.
How eMammal works
Contributors and partners set up camera traps to snap photos of passing animals. The images are uploaded and tagged through the eMammal app. Azure then processes the metadata and images, and machine learning identifies which images contain animals. The data and images can be used by scientists and the public on the eMammal site.
How eMammal uses Azure
- DevOps Pipelines for managing builds and deployment
- Blob Storage for image storage and access management
- Service Bus for communication among virtual machines
- Peered Virtual Networks for security boundary management
- ARM templates for deploying multiple virtual networks and virtual machines
- Virtual machines for running the eMammal Expert Review Tool, the database of image metadata and image processing services
eMammal’s Azure implementation was developed by Elastacloud.