TremorSense

A wearable, IMU-based system paired with an embedded machine learning model for real-time detection and severity classification of Parkinsonian and essential tremor. The current bench build runs entirely on an ESP32-C3, with no cloud connection, and streams a local dashboard of orientation and tremor band metrics.

Development is moving toward a dedicated smart wearable with improved calibration, quaternion based orientation estimation and on-device inference. The signal processing already forms the backbone of the lab's rehabilitation robot work. The near term goal is a low cost, deployable monitoring tool. Longer term it integrates into a rehabilitation robotic arm, extending from passive monitoring to active support for motor recovery after partial stroke.