Reliable, Real-Time Acceleration Measurement Toward Automated Monitoring of Driver Behavior

As in other countries, Canada needs to be ready to care for an aging population. Innovation is needed to make health care more efficient and effective. One solution is to monitor health remotely using the Internet. Driving is a high-order function, and its loss can be catastrophic for an elderly person’s independence. Unfortunately, physicians do not have the tools they need to adequately determine their patients’ driving ability. This thesis details the design and implementation of a remote patient monitoring system that is capable of real-time monitoring, tests its performance, and utilizes it to observe aspects of driver behavior. Data were collected on sixteen acceleration/deceleration profiles using accelerometers, GPS, and dashboard velocity. The resulting acceleration waveforms were filtered with an adaptive filtering algorithm and compared. Differences between hard and soft acceleration profiles are clearly visible, and notable features of each may be characteristic of individual driver behavior.