Developing and deploying a novel bluetooth tracking system utilising machine learning to understand pollinator foraging behaviour

Grantham Scholar Christopher Noroozi will look at developing a new method of bee tracking to better understand pollinator foraging behaviour.

The project

Critical to conservation is understanding how animals use landscapes, but we know very little about how central place foragers learn about & use their surroundings. Current pollinator tracking methods are limited in scale and accessibility. Thus, we will develop a novel method of bee tracking by deploying a series of low-energy bluetooth (BLE) transmitters across a landscape. Each one transmits, using a rotating high-gain antenna, the time, its identity and the angle. The receiver, attached to the bumblebee, will record these values, and the associated signal strength.

Through careful use of on-device machine learning, we will be able to minimise energy usage, allowing either a small battery, supercapacitor or thin-film solar-cell to power the chip. Our prototype weighs less than 40mg. The solar version will weigh less than 20mg.

Initially released at the nest, the bee will be recaptured upon its return to the nest and the data read. The recorded times, angles and identities of signals from the different transmitters can then be combined, using a Bayesian approach, to infer the flight path of the bee.


Dr Michael Smith

Department of Computer Science


Dr Michael Mangan

Department of Computer Science