Researchers developed a radar system to classify pollinating insects without capturing or killing them. This approach bypasses the lighting and weather limitations that plague traditional machine learning vision systems. By analyzing radar signatures, the tool enables non-invasive tracking. Practitioners can now monitor biodiversity in cluttered environments where cameras typically fail to capture usable images.