The MOLRA approach employs a fleet of semi-autonomous UAVs to collect audio, visual, and eDNA samples from a study site without ever needing to step foot in the site itself. Our surveys are supported by the global Map of Life biodiversity data and modeling engine, a suite of machine learning and AI models and innovative eDNA processing methods, and the combined knowledge and experience of a team that has been studying rainforests and biodiversity for decades.
Our approach employs a fleet of semi-autonomous UAVs to collect audio, visual, and eDNA samples from a study site without ever needing to step foot in the site itself. Our surveys are supported by the global Map of Life biodiversity data and modeling engine, a suite of machine learning and AI models and innovative eDNA processing methods, and the combined knowledge and experience of a team that has been studying rainforests and biodiversity for decades.
For any site in the entire world, our robust Map of Life modeling engine predicts what terrestrial vertebrate and tree species are expected to be found there. For the Americas and Southeast Asia, these predictions are based on tens of thousands of high-resolution species distribution models developed by our team, which will soon be available for Africa and, finally, the entire world.
Our autonomous vegetation mapping missions deliver true-color maps at 5cm resolution along with a set of indices. These maps are fed through a set of AI models that automatically detect tree crowns, flowering canopies, and other important vegetation characteristics, estimate tree crown diameters and heights, and support habitat classifications of the entire area. This data all serves to inform our subsequent sampling missions and provide important insights.
We customize our UAVs to collect eDNA samples from the air, water, and leaf surfaces throughout the study site. These customized rigs, developed by our team, are extremely cost efficient – one water eDNA sample costs less than $5 to collect. Our team has developed novel methods to process samples in the field. eDNA allows us to detect species that might otherwise be missed in our other sample collection modalities, particularly freshwater fish. Most of our eDNA sampling missions are fully autonomous
A combination of autonomous and remotely-controlled UAVs collect audiovisual samples in the study area guided by our vegetation and habitat mapping results. We collect visual samples from the canopy and the understory, targeting trees, epiphytes, and herbaceous plants, and we collect audio samples at dawn, dusk, and night to target birds, amphibians, and bats. These samples are fed into our AI models to produce preliminary species identifications.
All raw audiovisual samples along with AI-based identifications and processed eDNA samples with identifications are uploaded onto the MOLRA Dashboard where taxonomic experts can review, confirm, or edit species IDs. Taxonomic experts from anywhere in the world can log on to review samples and collaborate with other experts, greatly speeding up the ID process compared to traditional survey methods. The MOLRA Dashboard also provides a suite of insights for users to explore.
Linking all our technologies and data with collaborators and end users is the MOLRA Dashboard, a dedicated online platform for sharing survey results and insights and engaging with identification experts and the local community. The MOLRA dashboard facilitates rapid and collaborative review of our samples and their predicted IDs, culminating in a robust, quality-controlled final species list. Accompanying this species list is a suite of additional biodiversity insights drawn from the Map of Life global databases, including species conservation status and the broader regional and global significance of the study site.
Professor of Ecology and Evolutionary Biology, Yale University
Team Lead
Professor of Ecology, Evolution, and Natural Resources at Rutgers University
EDNA Team Lead
Senior Conservation Ecologist at the Field Museum of Natural History
EXPERT ENGAGEMENT AND IPLC LEAD
Associate Professor of Genetics at the Federal University of Amazonas
EDNA Team Lead
Interested in partnering or collaborating with us? Reach out to our team manager, Alex Killion, at [email protected]