Discipline: Computer Sciences and Information Management
Subcategory: Computer Science & Information Systems
Amari A. Vaughn - North Carolina A&T State University
The smartphones that people use daily, have multiple built-in sensors, each having a specific function which helps the device perform efficiently. These sensors can potentially be used to detect events or gather data about our human activities. The use of smartphone sensor data for machine learning could have numerous applications in finance, healthcare, entertainment, etc. In this research, we develop mechanisms to gather sensor data, and study how the data can be used to differentiate patterns in user physical motions. We developed a smartphone application for android platform to gather data from a mobile device. To validate findings in data, we applied sensor calibrating and/or filtering depending on the type of sensor data we want to analyze. To detect the device motion, our experiments will use the smartphone accelerometer and gyrometer. In this research, multiple behaviors were distinguished using sensor data. For example, using the smartphone accelerometer sensor, we found differences in data patterns from walking and standing. Data collected while the user is walking shows an average acceleration of 7.8 m/s^2 on for the y-axis of the device and is identical to the average acceleration when the user standing. To differentiate walking from standing, we analyzed unique changes in data such as, the x-axis acceleration for walking, which has values ranging between 14 m/s^2 and -12 m/s^2. The average x-axis values from standing shows little to no change in acceleration. The results of our tests show trends and/or characteristics in data which can be used to uniquely identify motions affecting the device. In conclusion, we can identify and differentiate human activities using sensor data. Future research involves the use machine learning to detect environmental changes and recognize human activities.
Funder Acknowledgement(s): This research was conducted as a part of NC A&T SU Summer Data Science and Analytics STEM Education Research Program and funded by the National Science Foundation (NSF) under Grant HRD#1623358.
Faculty Advisor: Mohd Anwar, manwar@ncat.edu
Role: Data Analysis, Data Collection, and Experiments.