The AMBER Project on Channel News Asia (CNA) Feature!
- liyuenchiew
- 12 minutes ago
- 2 min read
The Tropical Ecology and Entomology (TEE) Lab at the Asian School of the Environment, College of Science, Nanyang Technological University (NTU), Singapore, was recently featured on Channel News Asia (CNA) for its work on the AMBER Project — AI and Automated Monitoring of Insects (AMI) — showcasing how artificial intelligence is being integrated into biodiversity monitoring in Singapore.
Developed through a collaboration between the Alan Turing Institute, the UK Centre for Ecology & Hydrology (UKCEH), and the Aberdeen Charitable Foundation, the AMBER initiative represents a major step toward transforming how biodiversity is monitored globally. The project made its debut in Singapore on 24 November 2023, marking the first rollout of the system in a tropical region (see here for more details).
Our work was featured as part of the docu-series Raising Expectations which premiered on CNA, on the 15th February at 9.00pm, and explores the pros and cons of AI for the world today. The AMBER project features in Episode 3 : Creating A Healthier World? which aired on 17 February at 9.00pm, and showcases our work with UKCEH and Gardens by the Bay to develop automated technology to monitor nocturnal biodiversity. Don’t worry if you missed it – you can find it here at the CNA YouTube channel.
Why AI is needed for biodiversity monitoring?
Automated monitoring systems generate enormous amounts of image data. Every night, light traps capture thousands of images of insects attracted to the system. For taxonomists and entomologists to manually go through and identify every single image would be virtually impossible. This is where AI becomes essential.
Artificial intelligence allows us to rapidly analyze these vast datasets, automatically classifying organisms captured by the system. Rather than replacing experts, AI acts as a powerful assistant — sorting, filtering, and accelerating the identification process so researchers can focus on verification and deeper ecological analysis.
The Tropical Challenge
Applying AI in tropical ecosystems presents unique challenges. Tropical systems are hyper diverse. We have far more species co-occurring in the same habitats compared to temperate regions. At the same time, there is a shortage of taxonomists and regional experts who can identify these species. This creates a bottleneck: AI systems require well-labeled training datasets, but generating those datasets requires expert knowledge.
Most existing algorithms have been trained largely on temperate insect datasets from Europe and North America. As a result, their accuracy in tropical environments remains limited. At present, our preliminary data suggests that the system performs reasonably well at distinguishing broad taxonomic groups, but species-level classification remains a work in progress. The algorithm still requires substantial training using tropical data.
Building the Future of Tropical Biodiversity Monitoring
AI in tropical biodiversity research requires investment — not only in technology, but in people. We need more trained taxonomists, more local expertise, and more labeled tropical datasets to properly train these algorithms. The AMBER Project represents a step toward bridging that gap. By adapting automated monitoring tools for tropical systems, we are working toward a future where biodiversity data is more accessible, more standardized, and more globally representative.
Watch the full series here:
Learn more about the project here:
Comments