Developing a Novel Bluetooth Tracking System for Tracking Insects in the Landscape, Using Probabilistic Machine Learning

A bee’s foraging flight—zigzagging across fields, hedgerows, and gardens—remains almost impossible to track over distances greater than a few metres. That gap matters because wild pollinators are declining globally, and conservationists and farmers cannot see where insects actually go. Current tracking technology for larger animals—GPS collars, satellite tags—does not work for flying insects. The only alternative, a radar system called azimuth scanning harmonic radar, costs tens of thousands of pounds, requires flat open terrain, and is too cumbersome for most landscapes. Without a practical tracking method, scientists cannot answer basic questions about how bees move between flower patches, how far they travel, or which landscape features guide or block their routes. This project aims to build a lightweight, Bluetooth-based tracking system combined with probabilistic machine learning to follow insects across real, cluttered landscapes. If it works, researchers could for the first time map the foraging routes of bees and other flying insects at landscape scale. That would directly inform where to plant wildflower strips, how to design agricultural margins, and which habitats to protect—decisions that currently rely on guesswork. The system itself is fundamental science: it develops new methods for tracking tiny, fast-moving objects in complex environments, with potential future applications in environmental monitoring and precision agriculture.