Week 4 – Pedestrian Tracking with Shoe-Mounted Inertial Sensors
The authors propose a tracking technology for the real world applications of MR. Computer Vision algorithms are being looked at as the ultimate solution to the tracking problem but there are a lot of complexities involved with it and they are not mature enough to be successful alone. The other alternative is inertial sensors. He highlights way to overcome the problems of inertial sensors which can maintain accuracy of a few millimeters for one second by trying to circumvent the drift problem in the gyros, which cause a tilt error, by navigating the open loop for only for about 0.5 seconds at a time. The system makes use of a wireless inertial sensor which is small enough to fit into shoe laces and run on low power.
When a person walks, their feet alternate between a stationary stance phase and a stride phase. The Navshoe takes advantage of this and improves accuracy by applying zero velocity update (ZUPT) to EKF when it detects a stance phase after each stride to correct the drift velocity error. It can be combined with GPS to extend the range and correct yaw drift error. It is not affected by magnetic disturbance and works well wirelessly without any calibration indoors and outdoors. Combined with the fact that it can detect not only location but altitude and orientation accurately makes it quite impressive.
I would like to ask the author if it we can take steps to add this technology to the phone. It does not need too much battery and is light-weight, wireless and accurate, all the features desirable for mobile phone hardware.
Comments are closed.