SENSEable Shoes (proposal)

Abstract

SENSEable Shoes can understand human behaviors and thus provide a tangible wearable interface for interaction. By getting the values from twelve pressure sensors under shoe-pads in shoes (left foot and right foot), the algorithm can identify distinct feet gestures and predict the status of human behaviors related to shoes (such as changing the center of gravity of feet soles to the right/left/front/back, standing with two legs or one legs, sitting straight, sitting and leaning back, sitting and moving feet, shaking legs, walking, running, going upstairs or downstairs, jumping).

Problem

This is a Physiological Data Modeling problem whose purpose is to classify the person's behavior patterns based on the sensor measurements. In order to customize according to different person’s behavior patterns, the software will use python 2.7 with SciPy, Numpy, and Matplotlib API to implement a machine learning algorithm.

Midterm Milestone

The algorithm predicts static statuses of human behaviors related to shoes (standing with two legs or one legs, sitting straight, sitting and leaning back, changing the center of gravity of feet soles to the right/left/front/back).

Final Milestone

The algorithm predicts dynamic statuses of human behaviors related to shoes (sitting and moving feet, shaking legs, walking, running, going upstairs or downstairs, jumping). A tangible user interface will be constructed to interact with the shoes. (Sorry I haven’t come up with an idea about the interface.)

References

Chen, M., & Huang, B. (2008). Intelligent shoes for abnormal gait detection. 2008 IEEE International Conference on Robotics and Automation, 2019-2024. Ieee. doi:10.1109/ROBOT.2008.4543503