Browsing by Author "Yu, Xun"

Now showing 1 - 5 of 5
  • Thumbnail Image
    Item
    Integrated Approach for Nonintrusive Detection of Driver Drowsiness
    (Center for Transportation Studies, University of Minnesota, 2012-10) Yu, Xun
    This project is the extension of Northland Advanced Transportation System Research Laboratory (NATSRL) FY 2008 and FY2009 projects titled, “Real-time Nonintrusive Detection of Driver Drowsiness,” which aims to develop a real-time, nonintrusive driver drowsiness detection system to reduce drowsiness-cause accidents. In our previous research, nonintrusive sensors for drivers’ heart beat measurement were developed and implemented on the vehicle steering wheel. Heart rate variability (HRV) was analyzed from the heart beat pulse signals for the detection of driver drowsiness. Promising results were obtained. However, one of the major issues with the previous system is using only one parameter, Low-Frequency (LF)/High-frequency (HF) ratio of HRV, to access the driver's status, which has relative high variability and has different changing patterns for different drivers. In this project, we used multiple parameters for the drowsiness detection, including the LF/HF ratio, steering wheel motion variability and Electroencephalography (EEG) parameters. Correlations between these parameters are analyzed.
  • Thumbnail Image
    Item
    Intelligent Pavement for Traffic Flow Detection – Phase I
    (Intelligent Transportation Systems Institute, Center for Transportation Studies, University of Minnesota, 2012-09) Yu, Xun
    This project explored a new approach in detecting vehicles on a roadway by making a roadway section itself a traffic flow detector. Sections of a given roadway are paved with carbon-nanotube (CNT)/cement composites; the piezoresitive property of carbon nanotubes enables the composite to detect the traffic flow. Meanwhile, CNTs can also work as the reinforcement elements to improve the strength and toughness of the concrete pavement. In contrast to current traffic flow detection technologies that require separate devices to be installed either in the pavement or over the road, the proposed sensing approach enables the pavement itself to detect traffic flow parameters. Therefore, the proposed sensor is expected to have a long service life with little maintenance and wide-area detection capability.
  • Thumbnail Image
    Item
    Intelligent Pavement for Traffic Flow Detection – Phase II
    (Intelligent Transportation Systems Institute, Center for Transportation Studies, University of Minnesota, 2012-09) Yu, Xun
    This project is the extension of a Northland Advanced Transportation System Research Laboratory (NATSRL) FY09 project, titled as “Intelligent Pavement for Traffic Flow Detection”, which aims to explore a new approach in detecting vehicles on a roadway by making a roadway section as a traffic flow detector. Sections of a given roadway are paved with carbon-nanotube (CNT) enhanced pavement; the piezoresitive property of carbon nanotubes enables the pavement to detect the traffic flow. Meanwhile, CNTs can also work as reinforcement elements to improve the strength and toughness of the concrete pavement. The proposed sensor is expected to have a long service life with little maintenance and wide-area detection capability. In the FY09 project, lab tests demonstrated that CNT based cement composite can detect the mechanical stress levels for both static and dynamic loads. In the FY10 project, the research was extended to cement mortar, which has much higher mechanical strength and more useful in real applications. The effects of water level and CNT doping levels on the piezoresistivity of the composites were also studied. Preliminary road tests were performed for the evaluation of this new traffic sensor.
  • Thumbnail Image
    Item
    Real-time Nonintrusive Detection of Driver Drowsiness
    (Center for Transportation Studies, University of Minnesota, 2009-05) Yu, Xun
    Driver drowsiness is one of the major causes of serious traffic accidents, which makes this an area of great socioeconomic concern. Continuous monitoring of drivers’ drowsiness thus is of great importance to reduce drowsiness-caused accidents. This proposed research developed a real-time, nonintrusive driver drowsiness detection system by building biosensors on the automobile steering wheel and driver’s seat to measure driver’s heart beat signals. Heart rate variability (HRV), a physiological signal that has established links to waking/sleepiness stages, is analyzed from the heat beat pulse signals for the detection of driver drowsiness. The novel design of measuring heat beat signal from biosensors on the steering wheel means this drowsiness detection system has almost no annoyance to the drivers, and the use of a physiological signal can ensure the drowsiness detection accuracy.
  • Thumbnail Image
    Item
    Real-Time Nonintrusive Detection of Driver Drowsiness – Phase II
    (Center for Transportation Studies, University of Minnesota, 2010-12) Yu, Xun
    This project is the extension of the Northland Advanced Transportation System Research Laboratory (NATSRL) FY 2008 project titled “Real-time Nonintrusive Detection of Driver Drowsiness,” which aims to develop a real-time, nonintrusive driver drowsiness detection system to reduce drowsiness-caused accidents. Biosensor is built on the vehicle steering wheel to measure driver’s heartbeat signals. Heart rate variability (HRV), a physiological signal that has established links to waking/sleepiness stages, thus can be analyzed from the pulse signals for the detection of driver drowsiness. The novel design of measuring heartbeat signals from biosensors on the steering wheel and seatback makes this drowsiness detection system one with almost no annoyance to the driver, and the use of this physiological signal can ensure the accuracy of drowsiness detection. In Phase I, a biosensor with a pair of electrodes built on steering wheel was tested for the measurement of heartbeat for HRV analysis. However, this design requires the driver put both hands on the steering wheel to measure the heart rate. In Phase II, a new biosensor is designed that can measure heart rate even when only one hand is on the steering wheel, which happens very often in real driving situations. More extensive lab tests were carried out to study the change of HRV signals with driver drowsiness.