Accelerometer Lab

The Accelerometer Lab was developed with Sintesilabs as part of an exhibition in collaboration with the Max Planck Institute of Animal Behavior of Berlin. The exhibition explores how scientists use tracking technologies—such as accelerometers and camera traps—to study animal behavior in the wild and better understand movement patterns and activity.

My contribution focused on designing and building an interactive laboratory station that allows visitors to explore how accelerometer data works and how it can be interpreted to understand motion.

The installation is based on a micro:bit device equipped with accelerometer, which visitors can move, rotate or shake while observing how the sensor data changes in real time. The system collects motion data and processes it through a Raspberry Pi-based setup, where it is visualized and analyzed.

To go beyond raw sensor readings, the project experiments with machine learning models built with TensorFlow, attempting to recognize patterns in the accelerometer signals and classify different types of movements. This reflects the same type of analysis used in scientific research, where accelerometer data is often used to infer behavioral states from movement patterns.

The goal of the installation is educational: to make visitors understand how a small sensor measuring acceleration on three axes can become a powerful tool for studying motion, behavior and activity, both in animals and humans.

The project combines embedded hardware, sensor data acquisition, real-time visualization and machine learning, translating complex scientific concepts into a tangible interactive experience for the public.

Technologies

• AI
• Machine learning
• micro:bit
• Raspberry PI
• Tensorflow

Links

• Descrizione in italiano
• Tech details on the accelerometer lab