Jacob Kamminga's Thesis Details a Simple Accelerometer-Based Tracker for Wildlife Protection

University of Twente researcher Jacob Kamminga has developed a new tool to help guard animals against poaching — using nothing more than a simple accelerometer linked to a wide-area communications network, plus a machine learning system trained on only a small amount of labeled data.

"Animal activity recognition (AAR) is a new field of research that supports various goals, including the conservation of endangered species and the well-being of livestock," Kamminga writes in his thesis on the topic. "Over the last decades, the advent of small, lightweight, and low-power electronics has made it possible to attach unobtrusive sensors to animals that can measure a wide range of aspects such as location, temperature, and activity. These aspects are highly informative properties for numerous application domains, including wildlife monitoring, anti-poaching, and livestock management."

"In this thesis, we focus on AAR that aims to automatically recognize the activity from motion data – on the animal – while the activities are performed (online). Specifically, we use motion data recorded through an inertial measurement unit (IMU) that comprises an accelerometer, gyroscope, and magnetometer to classify up to eleven different activities."

Kamminga's research centred around the use of an off-the-shelf Inertia ProMove-Mini intertial measurement unit, which includes two three-axis accelerometers, a three-axis gyroscope, a three-axis magnetometer, temperature sensor, and barometric pressure sensor. These sensors were attached to livestock and their movement patterns recorded — then a small portion of labelled data used to train a machine learning system to recognize different types of movement.

Kamminga found that it was sufficient to use the data from only one sensor — the accelerometer — to accurately estimate what an animal is doing at any given time, including whether it is potentially fleeing from poachers. By dropping down to a single sensor, and only transmitting data when an abnormal event is detected, Kamminga claims it would be possible to monitor animals in the wild in order to protect them from poaching.

"Linking wild animal movement recorded using sensors with remotely sensed imagery and GIS [geographic information system] models is promising technology to better understand the ecological requirements of species," says Professor Andrew Skidmore of Kamminga's thesis, "as well as inform management and policy decisions with conservation outcomes and biodiversity."

The thesis is available for download under open-access terms from the University of Twente. Those working on similar initiatives are invited to participate in ElephantEdge, a competition to build the world's most advanced wildlife tracker using Edge Impulse and the Avnet IoT Dashboard — and with up to $5,000 in prizes available for the 10 winning submissions.