Motion-Artifact Resistant Design of Photoplethysmograph Ring Sensor for Driver Monitoring

Through extensive literature search, it has been found that there exists strong correlation between the Mental Workload of an automobile driver with his or her physiological measurements, especially the Heart Rate Variability (HRV) measurements. In order to make a driver's HRV monitoring on the road possible, a new photoplethysmograph (PPG) Ring Sensor prototype has been devised specifically for countering the problem of motion artifact that almost all wearable PPG sensors are facing. Most importantly, the motion artifacts on PPG signals caused by motion in the direction of blood flow in the digital arteries along the finger flanks have been eliminated using a customized Adaptive Noise Cancellation algorithm. Both the lab results and the road test results have suggested that the new Ring Sensor is indeed capable of rejecting motion artifacts in all three possible motion axes and producing considerable amount of usable beat-to-beat heart rate data on the road for HRV analysis. The problem of occasional missing data on the road has also been tackled with a suitable linear curve-fitting algorithm. Also, the sunlight saturation problem is dealt with using a simple DC averaging reference circuit. The final road test has proven the validity of the driver Mental Workload model and the validity of the Ring Sensor in monitoring the HRV of the driver on the road. Besides the application on driver monitoring, the Ring Sensor can also be used for other forms of wearable monitoring such as jogging. Thesis Supervisor: Haruhiko H. Asada Title: Professor of Mechanical Engineering