Silent Pollinator Tracking With mmWave Radar

https://spectrum.ieee.org/mmwave-radar-insects-pollinators

Publish Date: 2026-05-23 09:00:01

Recent advancements in radar technology may offer a new way to monitor bees and other pollinating insects without invasive techniques. Researchers from Europe have developed a radar system that leverages micro-Doppler signatures—small, time-varying patterns in radar reflections caused by insect wingbeats—to distinguish between different species of pollinators. By using millimeter waves, which match insect sizes better than other forms of radar, the team trained a machine learning model to analyze over 70 features from radar reflections. The model showed an 85 percent accuracy in species identification and up to 96 percent when distinguishing between broad insect families. This method, which does not harm the insects, could one day lead to a global database for the instant classification of any insect, while also potentially aiding in the monitoring of pest behaviors.

Key Points:
– Researchers are developing a radar system to monitor pollinators using micro-Doppler signatures from millimeter waves.
– The machine learning model distinguished species with 85 percent accuracy and family-level distinctions with 96 percent accuracy.
– The radar method is non-invasive and utilizes low power levels to avoid harming insects.
– Researchers aim to create a portable version of the technology and a global database for classification.
– Future applications could include monitoring pest behaviors and tracking invasive species.