Oral Candidacy - Rachael Purta
|Start:||8/11/2016 at 10:00AM|
|End:||8/11/2016 at 1:00PM|
|Location:||258 Fitzpatrick Hall|
Faculty and students are welcome to attend the presentation portion of the proposal defense.
August 11th 258 Fitzpatrick 10:00 am
Adviser: Dr. Striegel
Dr. Christian Poellabauer Dr. Dong Wang Dr. Ron Metoyer
Dr. David Hachen
Bluetooth Low Energy Proximity Detection Techniques for Measuring Social Interaction and Health Behaviors
The study of social interaction has benefited greatly from recent technology, especially mobile phones and their various sensors. There are still many types of socially-relevant data, however, that cannot be measured by mobile phones alone. To measure these types, sociology and psychology research on social interaction usually relies on surveys, which can be biased and incomplete. This work proposes using proximity-based techniques, especially those based on Bluetooth low energy, to measure several socially-relevant case studies, including exercising with friends, food choices and eating with friends, and interactions between children and their caretakers. Bluetooth low energy (BLE) techniques that are proposed are using BLE peripherals (Fitbit) for proximity measurement/enhancement, using stationary BLE beacons for proximity for mobile phones, and using moving BLE beacons and stationary scanners for pseudo-positioning (using proximity to estimate position). BLE beacons are shown to be useful for measuring proximity for close distances (up to 2 meters) on an iOS phone, but beacon readings from a Raspberry Pi are shown to need special filtering techniques to achieve similar accuracy. Various ways of improving distance estimation, as well as other experiments necessary to better understand BLE for more general social sensing applications, are also discussed.