Capgemini’s New AI Model Helps Protect Biodiversity By Listening To Insects

Data and artificial intelligence experts from Capgemini, in collaboration with Naturalis Biodiversity Center and Amazon Web Services, have developed an AI model that can help protect biodiversity by automating and improving the identification of insects through sound recognition.

Currently, identifying insect species by their sounds is a challenging task done by only a few experts. Targeted AI solutions developed by Capgemini can enable scientists to analyse terabytes of information within 24 hours, the company said.

The AI solution was developed as part of Capgemini’s sixth Global Data Science Challenge for a sustainable future, an annual initiative whereby employees are given the opportunity to harness their data and AI expertise to solve a real-world issue.

For this year’s challenge, teams from Capgemini collaborated with Naturalis—the National Research Institute for Biodiversity based in the Netherlands—for their expertise in the field, and with AWS for their scalable machine learning platform, to tackle the issue of insect monitoring. The winning team developed an AI-based acoustic model that can identify different insect species with almost 92% accuracy.

“Listening to and identifying insect species is critical to preserving them. I’m inspired to see how our colleagues have collaborated to apply the power of AI to conserve our natural ecosystem and help protect the future of our planet,” said Niraj Parihar, CEO of Insights and Data Global Business Line at Capgemini.

Insects play a critical role in the global ecosystem, for example, as pollinators, indicators of an ecosystem’s health and food source for larger animals. However, the insect population is increasingly being impacted by changes in land use, pesticides, urban development and shifts in climate.

Elaine van Ommen Kloeke, programme manager at Naturalis, said, “If we lose insects, it would be a devastation felt not just by the species that directly rely on insects as a food source, but everyone in the ecosystem—including humans. It could lead to the loss of about 35% of the world’s food crops. With this initiative, we aim to make this technology accessible to research communities worldwide to support with the identification of insects in any given environment.”

Insects are also incredibly hard to monitor and, given that there are millions of species, it cannot be done reliably by humans. Although image recognition works for some species, it cannot successfully detect species that are inconspicuous to the human eye. It is here that acoustic recognition plays a role.

Together with traditional and camera-based monitoring, acoustic recognition will allow for identification of insect species that would otherwise go undetected, enabling scientists to monitor and conserve vital populations.

Tanuja Randery, managing director at AWS—Europe, Middle East and Africa, said, “We’ve seen over the years how data and AI can contribute to building a better, more sustainable future, and it’s fantastic to see that come to life.”