Estimating Energy Consumption using Inertial Sensing on a Smartphone

Description or Abstract of the Project: 

This project draws from the "Energy Consumption Dataset," collected from experiments conducted at the USC Division of Biokinesiology. This study aimed to explore whether probabilistic techniques could be used to improve predictions and whether alternative sensors like gyroscopes could be used to predict energy expenditure.The goal of this project was to determine whether on-body inertial sensing with mobile phones together with probabilistic machine learning techniques could be used to characterize energy expenditure as measured by rate of oxygen consumption (VO2, mL/min). Experiments were conducted for one type of activity: steady-state treadmill walking. Steady-state treadmill walking was chosen because of the ease of data collection over a range of walking speeds. At the time of this study, existing mobile phones did not have the required sensor suite or sampling capability as was desired; hence, an on-body inertial sensor interfaced to a laptop or a mobile phone was used. The above information is from CENS annual report PART 05.

Principal Investigator (Last Name, First Name): 
Sukhatme, Gaurav
Email Address of Principal Investigator:
Persons in Charge of Data Collection (Last Name, First Name): 
Emken, Adar; Vathsangam, Harshvardhan
Email Address of Persons in Charge of Data Collection:;
Data Registered by (Last Name, First Name): 
Mathieu, Camille
Email Address for Person Registering the Data: 
Institution and Department: 
UCLA CENS; USC Department of Computer Science, Viterbi School of Engineering
Date of Collection : 
Monday, April 19, 2010 to Friday, June 18, 2010
Data Collection Methods: 
Eight healthy adults (four men, four women) participated in the study. Each participant was asked to walk at 5 speeds (2.5, 2.8, 3.0, 3.3, and 3.5 mph) on a motorized treadmill for 7 minutes of recording time per speed. Following a transition between speeds, VO2 readings were allowed to stabilize for 2 minutes prior to data collection.
Equipment Used in Data Collection: 
Sparkfun 6DoF Inertial Measurement Unit, Freescale MMA7260Q tri-axial accelerometer, 2 Invensense IDG300 gyroscopes, RN-41 Bluetooth module, HTC Nexus one running Android 2.2.1, MedGraphics Cardio II metabolic system with BreezeSuite v6.1B (Medical Graphics Corporation).
Description of Data Type: 
Data Format/File Type (example: jpg, xls, txt). Do NOT put periods in the format type or you will get an error message.: 
Are These Data Shareable?: 
Keywords : 
Energy consumption
Date of Registration: 
Thursday, March 6, 2014