Naturalis and the Arise initiative needed a practical way to interpret large volumes of insect acoustic data so conservation teams could track biodiversity change and identify species more effectively.
Reported impact
Naturalis improved insect-species identification performance from 76% to 92%, published the challenge results for others to build on, and established a scalable starting point for automated acoustic biodiversity monitoring.
Naturalis worked with Capgemini and AWS through the 2023 Global Data Science Challenge to turn raw insect audio into an AI-assisted biodiversity monitoring workflow. The customer contributed biodiversity context and acoustic recordings from the Arise program, then challenge teams used Amazon SageMaker machine-learning services to develop and test species-classification models against five hours of recordings covering 66 insect species. The winning model identifies species from sound with about 92% accuracy, improving prior performance while giving conservation researchers a practical starting point for automated monitoring. Capgemini used the challenge to orchestrate experimentation, published the results afterward, and is continuing development with Naturalis and AWS so other biodiversity organizations can reuse and extend the approach.
Naturalis supplied insect audio recordings and biodiversity context from the Arise initiative, teams trained and tested models in Amazon SageMaker, and the winning AI model outputs species classifications that Naturalis and partner organizations can use to monitor insect populations more efficiently.
"By learning from five hours of recordings of 66 different insect species, each team attempted to build an AI that can identify which species could be heard."
"By the end of the competition, there was an improvement in performance*(macro averaged F1-Score) from 76% to 92%."
"Capgemini collaborated with Naturalis, the National Research Institute for Biodiversity based in the Netherlands, and with Amazon Web Services (AWS) to launch the latest edition of the Global Data Science Challenge (GDSC) and create a new AI model that can identify insects with ~92%* accuracy."
"Following the GDSC’s conclusion, the entirety of the results have been published, enabling other organizations and teams to pick up where the challenge has left off."
Review how this reference was scored and trace each normalized claim back to its documented source.
Scoring methodology
Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.
Traceability
Field coverage shows which facts each source supports. Dates are shown only when documented by the source.
Related
Capgemini Sample Jun 9, 2026 Federal Ministry of Digitalization and State Modernization (BMDS) BMDS needed to reduce long, complex planning and approval procedures for infrastructure projects that routinely exceeded legal timelines and burdened case workers with documentation-heavy manual workflows. more than six months beyond the legally defined limit Capgemini Sample Dec 17, 2025 Amsterdam City Archives Amsterdam City Archives needed to make a massive paper archive usable for the public despite handwritten documents, outdated Dutch, and limited search tools that made discovery difficult. 50 kilometers of paper archives AI Search · Chat with History · Handwritten Character Recognition (HCR) Capgemini Sample Dec 5, 2025 German Federal Employment Agency The BA needed a privacy-compliant way to reduce the manual effort of turning Requests for Change and user stories into high-quality Jira tickets under increasing workload and limited staff capacity. Aleph Alpha · CrewAI · Jira 