LAUSR

Abstract The paper proposed a spot vehicle speed detection method bases on a smart infrastructure called Smart Road Stud (SRS), which has functionalities of vehicle detection, wireless transmission, self-powering module and a controlled light-based traffic guidance as a compact road safety stud. Traditional dual-node speed detection system deployed two separated nodes with relative several meters long gap. The SRS integrates two identical three-axis geomagnetic sensors with a short pitch less than 10 cm on a single board to form an integrated device for speed detection, which result the low time resolution of sample points with high driving speed. The slight delay ununiformity of two raw detection waveforms can induce a big error in time difference calculation. The paper proposed standardized processing and Local-Optimization-Fitting (LOF) method for speed calculation, for solving the problem of low time signal resolution and waveform ununiformity induced by different electrical characteristics of two channels. Field testing results show the LOF method for spot vehicle speed estimation is over 93% accuracy, which is worthy of practical application.