NEA is deploying machine learning and thermal cameras to track rats across Singapore
Source: CNA Tech
Singapore's National Environment Agency is rolling out thermal cameras and machine learning models to monitor rat activity, cutting manual review time and expanding surveillance coverage across hotspots like Little India and Geylang.
The National Environment Agency is getting serious about tech-enabled pest control. NEA has deployed 40 thermal cameras across Singapore between January and April this year, using machine learning to automatically identify rat movement in footage — a task that previously required hours of manual review by enforcement officers. The agency's rat control team is also using passive infrared cameras to study rodent behaviour and movement patterns, which has revealed that rats commonly use external AC units and pipes to access upper floors of buildings in search of food.
Data from the surveillance push shows results: the average number of rat burrows detected per two-monthly cycle dropped from 5,400 in early 2025 to around 4,200 in the second half. NEA and the Singapore Food Agency took about 1,300 enforcement actions against premise owners for rat-related lapses last year, up from just over 1,000 in 2024, with roughly half tied to poor refuse management.
Why it matters for Singapore: This is a textbook example of a Singapore government agency applying AI and automation to a stubborn public health problem. It's not flashy, but the combination of thermal surveillance and ML-based analytics is delivering measurable outcomes — fewer burrows, better-targeted enforcement, and more efficient use of limited manpower.