Ultimately, the team was able to develop a simple prototype of a smart knee brace using relatively few components. With just two WICED Senses,a commercial off the shelf knee brace (with pockets sown on), and custom Android and web Apps, the full emBRACE system was developed. The WICED Senses synced to the Android App over a Bluetooth Low Energy communication line, using the mobile phone as a gateway to write the accelerometer and gyroscope data to the cloud. This data was then preprocessed using a combination of a low pass filter and a kalman filter to smooth it out and make it easier to perform machine learning. A multiclass SVM separated the data into four states: Standing up, with the knee extended at 180 degrees, sitting down with the legs parallel to the floor and the knee extended 180 degrees, sitting down with the legs forming a right angle and the knee at 90 degrees, and sitting down with the legs stretched out at 45 degrees with relation to the ground, and the knees extended straight out. These four states allow the user to determine how much he or she can extend his knee, which is very important to understand when assessing quality of recovery. These states were then written back to dynamoDB for retrieval and display by the web application.
There were multiple facets to completing this project: from high quantity data collection, data processing, machine learning, mobile app development, web application development, AWS interaction, and an in depth systems integration to combine the different components together.
There were multiple facets to completing this project: from high quantity data collection, data processing, machine learning, mobile app development, web application development, AWS interaction, and an in depth systems integration to combine the different components together.