Parkinson's disease and other neurodegenerative disorders cause subtle changes in motor function that are nearly impossible to monitor outside clinical settings. Existing solutions — wearables and camera systems — are intrusive and alter natural behavior. Our team built a compact wireless sensing platform that turns everyday dining objects into motion tracking devices, letting researchers capture real movement data during routine activity.
The device is a detachable sensor module housing a 6-DOF IMU, rechargeable battery, and Bluetooth Low Energy radio. It clips onto mugs, forks, spoons, and knives — objects users already interact with naturally — so data collection is unobtrusive by design.
I led the mechanical design, prototyping, and system integration. Early development explored straps, clamps, magnets, and adhesives as mounting strategies. Through iteration, we converged on a modular architecture: a self-contained sensor package paired with interchangeable mounts. The final mechanism uses a magnetically assisted twist-lock — secure enough to survive active use, simple enough for users with limited dexterity.
The central engineering tension was compactness vs. signal fidelity. The housing needed to rigidly couple the electronics to the host object so vibrations and motion transmitted accurately to the IMU — any flex would corrupt the tremor signal. The final prototype is 28 mm in diameter, weighs 13 g, and mounts reliably to 6+ dining objects.
We also built the full data pipeline: Arduino firmware for sensor acquisition, BLE for wireless transmission, and Python-based analysis tools. Frequency-domain processing cleanly separated voluntary movement from tremor-like motion, validating the system as a low-cost research platform for real-world motor monitoring.
