Title

Real Time Activity Recognition of Treadmill Usage via Machine Learning

Authors

Nathan Blank, Rose-Hulman Institute of Technology
Matt Buckner, Rose-Hulman Institute of TechnologyFollow
Christian Owen, Rose-Hulman Institute of TechnologyFollow
Anna Scott, Rose-Hulman Institute of TechnologyFollow

Document Type

Article

Publication Date

8-27-2016

First Advisor

Carlotta Berry

Second Advisor

Valerie Galuzzi

Third Advisor

Yosi Shibberu

Abstract

Our objective is to provide real-time classification of treadmill usage patterns based on accelerometer and magnetometer measurements. We collected data from treadmills in the Rose-Hulman Student Recreation Center (SRC) using Shimmer3 sensor units. We identified useful data features and classifiers for predicting treadmill usage patterns. We also prototyped a proof of concept wireless, real-time classification system.

Comments

RHURP 16-04

Recommended Citation

Blank, Nathan; Buckner, Matt; Owen, Christian; and Scott, Anna, "Real Time Activity Recognition of Treadmill Usage via Machine Learning" (2016). Rose-Hulman Undergraduate Research Publications. 13.
https://scholar.rose-hulman.edu/undergrad_research_pubs/13