Project: #47 Non-Intrusive Driver Distraction Monitoring Using Vehicle Vibration Sensing Progress Report - Reporting Period Ending: Sept. 30, 2017 Principal Investigator: Hae Young Noh Status: Active Start Date: Jan. 1, 2017 End Date: Aug. 31, 2018 Research Type: Basic Grant Type: Research Grant Program: UTC FAST Act Grant Cycle: 2017 TSET UTC Progress Report (Last Updated: Oct. 5, 2017, 10:02 a.m.) % Project Completed to Date: 30 % Grant Award Expended: 0 % Match Expended & Document: 0 USDOT Requirements Anticipated Research Outcomes We introduced a system of inertial sensors in a car seat to provide ambient monitoring of a driver's heart rate and heart rate variability (specifically, RR intervals). At every beat, the heart is polarized and depolarized to trigger its contraction, electrical activity which is often measured by an ECG. The R interval describes the depolarization of the main mass of the ventricle, causing the largest peak in an ECG. The RR interval is defined as the distance between the peaks of two R waves. The RR-interval, then, describes the duration of one complete cardiac cycle. In this project, we use inertial sensors instead of ECG, so the peaks of the waves we measure correspond to the heart's movement, not its electrical activity. The RR interval can be used to calculate heart rate and heart rate variability. We have been focusing on acquiring successive RR intervals of drivers from a car seat in noisy driving scenarios. The outcomes include our hardware system integrated with algorithm software that removes noise from car and driver movements and extracts RR-interval. Filtering, wavelet signal decomposition and extreme value analysis techniques are used for effective separation of heart rate information from other noise. The system has been tested real cars and drivers when the car is on and off. We plan to further expand the evaluation to various driving scenarios in the rest of the project period. Based on the developed system, we have demoed and presented posters at the 2017 Cyber-Physical Systems Week held in Pittsburgh. We also utilized our system as a data collection platform in the graduate level project courses that the PIs teach – Sensing and Data Mining in Smart Structures and Systems and Mobile and Pervasive Computing. These project courses promote the students’ interest in projects related to safe transportation systems. Anticipated Impacts Continuous heart rate and heart rate variability monitoring in cars can allow for continuous health monitoring of a driver (both physical and mental) which is critical for safe driving. Heart rate variability monitoring can be used to help track stress and fatigue, which are driver safety issues. Stressful life events have a significant effect on traffic accidents. Studies show that the impact of stress on professional driver performance and concluded that it is a direct risk factor for driver accidents. Drowsy driving is also risky. According to the National Highway Traffic Safety Administration (NHTSA), about 56,000 crashes are caused by drowsy drivers each year in the United States of America, which results in about 1,550 fatalities and 40,000 nonfatal injuries annually. Continuous heart monitoring of drivers in various driving scenarios allow timely detection and prediction of critical driver states related to their safety and potentially reduce accidents by providing proper guidance to the driver (e.g., alarm, haptic feedback, climate control, etc.) Did research results confirm or change practice? This research confirms our initial hypothesis that the drivers’ physiological status can be monitored using inertial sensors in the car, which in turn can be used for drivers’ stress and attention level. In particular, we focused on heart rate and heart rate variability monitoring, and our results so far present high accuracy of heart rate estimation (within 1 bpm). This research will change the current practice of intrusive and/or inconvenient heart monitoring, such as ECG, camera, wearables. Web Links N/A Issues The remaining challenges to be addressed in the rest of the project period include 1. validating the system performance under various driving scenarios, including different speed, terrain conditions, city vs. freeway driving, etc. 2. testing the drivers of various physical conditions, such as age, gender, height, weight. Accomplishments 1. Two PhD students (both female) have been supported by this funding. 2. The technology was demoed at the 2017 Cyber-Physical Systems Week held in Pittsburgh 3. A demo abstract is published at the 2017 Cyber-Physical Systems Week 4. The technology has been pitched to companies, such as Intel, Google to promote industry collaborations and potential commercialization. 5. The work is presented through multiple invited seminars at Stanford, CalTech, Princeton, Georgia Tech, etc. to promote collaboration and expansion of the work. 6. The developed hardware has been used as a data acquisition platform in graduate level project courses at CMU to promote student interest in safe transportation systems. 7. We are currently preparing a submission of a paper to 2018 HotMobile workshop.