The Project

It has become relatively easy to track someones physical activity. Pedometers, and heart rate monitors for example have been around for ages doing just this. More recently, products such as the Fitbit and Jawbone have provided amounts of data not previously available. However, outside of activity tracking, there is very little data on human movement itself on large scales.

The goal of project, is to develop a mesh network of sensors that an individual would wear on their feet, and have it collect data on that person's gait. Gait, is the pattern of movement of an individual's limbs. This data would then be analyzed by the system, and effectively monitor how a person walks. The system would present the data in a manner which is human readable. Ideally, a health professional would use this and be able to determine if an individual needs to increase the size of their stride, or even go as far as finding issues with an individual's muscles and bones in their lower extremities, before a problem emerges.

Project Status and Future Work

The team has met the core requirements of the project. We have develped a system, that collects data on a persons gait. It collects values such as force on various locations of the foot, acceleration, and orientation of the foot. It collects this data every 8 milliseconds. This timing, gives researches and medical professionals the ability to read data effectively. It also allows them to analyze running, and not just walking. The team also developed a mobile application that can control the wearable device. Data from the wearable device can be sent up to the mobile device, and compressed into a .csv file. That file, can then be sent to a web server over a WiFi connection. Below is an example of data collected where the black line is a healthy patient taking steps forward, and then the orange line is an unhealthy patient taking steps.

Future work on this project revolves around statistical analysis. A goal is to do real time statistical analysis and machine learning on the wearable device. This leads to a number of optimizations. Being able to tell what the user is doing (i.e. walking or sitting) and shutting down depending on that activity, and lead to significant storage and battery efficiency. Lastly, doing real time statistical analysis on the web server, for instand results can only improve the usefulness of the product to researchers and medical professionals.