An early warning system for traffic and road safety hazards using collaborative crowd sourcing

Vehicular Perception Based on Inertial Sensing: A Structured Mapping of Approaches and Methods

“…Jamakhandi and Srinivasa [30] detected humps, pits and abrupt elevation. Bello-Salau et al [10], Gunawan and Soewito [26], Ghadge et al [23], Kumar et al [36], Li and Goldberg [39], Pooja and Hariharan [53] and Rishiwal and Khan [55] identified bumps and potholes, with [23] validating the threshold with a K-means clustering and random forest. Gawad et al [22] trained a perceptron to generate the threshold, used to identify the presence of road anomalies.…”

“…8, although other domains such as frequency and time-frequency have also been explored. Statistical approaches such as Gaussian [53], Skewness [4], Kurtosis [4], median [4,38], root mean square (RMS) [31,39,61], moving average/average [4,5,12,38,51,59,62], moving variance/variance [4,5] and standard deviation [5,10,28,40,51,54] were used in the time domain to smooth the signals and extract features in the vibration data. Among the papers that used signal preprocessing, about 60% of them used statistical techniques.…”

“…Some studies analyzed in this review use vehicle perceptions developed in a final application. The papers that use sensors embedded in mobile devices, for producing data with less reliability, are generally used in crowdsourcing solutions with opportunistic sensing [1,23,32,36,39,40,53,55,59,62,65,67]. In this application type, the solutions georeference the transient and persistent events, sending them to a central server.…”

“…In this application type, the solutions georeference the transient and persistent events, sending them to a central server. In this server, a post-processing step is applied, where data from different sources are integrated to improve reliability, as in Gawad et al [22] and Pooja and Hariharan [53]. Subsequently, these data are made available back to users in applications such as advanced driver assistant system (ADAS) [1,2,48].…”

“…Euler angles Linear interpolation [40] Standard deviation Android library N/A [41] N/A N/A N/A [45] Low-pass zero-phase N/A N/A [48] N/A Rotation matrix N/A [49] Band-pass third-order Butterworth zero-phase N/A N/A [50] N/A Euler angles N/A [52] High-pass 3rd order Butterworth N/A N/A [53] Gaussian low pass N/A N/A [54] Standard deviation N/A N/A [55] N/A N/A N/A [59] Moving average N/A N/A [60] N/A N/A N/A [62] Low pass High pass Simple moving average (SMA) Lifting wavelet transform (LWT)…”