Building on insights from over 50 stakeholder interviews conducted by the team before I joined, we identified a core opportunity: a wearable glove that could track hand movement data for labor-intensive tasks. I led research into system integration for the MVP, consulting with experts including PhD students from Northwestern’s HaND Robotics Lab to determine the most effective sensing technology and visualization application. After evaluating various options, we selected Inertial Measurement Units (IMUs) for their balance of precision and integration flexibility. I prototyped an initial system using two IMUs, which transmitted data to a custom MATLAB script that visually interpreted the raw sensor output for hand motion analysis. 

With a better understanding of quaternion and matrix outputs from IMUs, we transitioned to a fully wearable prototype. We integrated the electrical components into a glove form factor by soldering and sewing the sensors, along with a compact battery and charging circuit for wireless operation. The data was fed into a new Python pipeline I developed, using PyBullet to simulate a 3D model of the hand. We tested several sensor configurations, including 1-, 2-, and 3-IMU setups. After extensive experimentation, we optimized the glove around a single IMU, using inverse kinematics algorithms to estimate full-hand motion from a single input point.