Stress and addictive behavior (e.g. smoking) lead to or worsen diseases of slow accumulation such as heart diseases and cancer. While traditional diseases caused by malnutrition or poor hygiene are becoming rarer, stress and addictive behavior continue to be widespread. Reliable inference of stress and addictive behavior using unobtrusively wearable non-invasive sensors in the natural environment of an individual still remains a formidable challenge due to non-specificity of the measures such sensors collect. In the AutoSense project, we have developed a comprehensive suite of wearable sensors that can be worn in the mobile environment to collect multiple physiological indices of stress and addictive behavior (e.g., ECG, Respiration, Alcohol, etc.). AutoSense is complemented by a software framework on the mobile phone called FieldStream that collects physiological measurements from AutoSense sensors, processes them to derive behavioral inferences, and uses these behavioral events to solicit self-reports on the phone, all in real-time. The entire end-to-end system has been worn by 50+ human volunteers for 2,000+ hours in their natural environments as part of various scientific user studies. From these real-life sensor measurements, we have developed robust models to infer psychological stress and to detect conversation episodes. We find that people are stressed 27% of their day and that the average duration of a conversation is 3.8 minutes, among several other interesting results on naturally occurring human behaviors. In this talk, I will introduce the AutoSense and FieldStream platforms, models for detection of stress and conversation, and describe the advances we are making in inferring addictive behaviors from sensor measurements collected in the natural environment.
Santosh Kumar is an Associate Professor of Computer Science at the University of Memphis, where he received an Early Career Research Award from the College of Arts and Sciences in 2008. He received his Ph.D. in Computer Science and Engineering from the Ohio State University in 2006, where his dissertation work won the SBC Presidential Fellowship award. In 2010, the Popular Science magazine named him one of America’s top ten brilliant scientists under the age of 38 for leading the development of the AutoWitness burglar tracking system and the AutoSense wearable sensor system. On the theory side, he is known for establishing new models of coverage with wireless sensors such as barrier coverage for intrusion detection and trap coverage for target tracking. More information on him is available at his homepage: http://www.cs.memphis.edu/~santosh/.