Road Safety Features Identification Using the Inertial Measurement Unit

• Published in: IEEE sensors letters Vol. 2; no. 4; pp. 1 - 4
• Main Authors: Baldini, Gianmarco, Giuliani, Raimondo, Dimc, Franc
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01.12.2018; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Accelerometers; Automotive engineering; classifier design and evaluation; Feature extraction; Gyroscopes; Identification; Inertial platforms; Irregularities; mobile environments; Road maintenance; Road safety; Sensor applications; Sensor systems and applications; Traffic accidents & safety; Traffic safety; transportation; Transportation engineering; Warping
• ISSN: 2475-1472; 2475-1472
• DOI: 10.1109/LSENS.2018.2880118

Recent studies have investigated the use of inertial measurement units to detect and report the presence of road irregularities on the surface and thus improve safety and maintenance of the transportation infrastructure. The results from the literature show that it is possible to detect the presence of various irregularities in the road surface and identify the type of irregularity with good accuracy using various statistical features. This article addresses specific irregularities called road safety features (RSF) and focuses on the identification of the RSF instances themselves rather than the RSF type identification, which leads to a more challenging problem. The authors have collected data from accelerometers and gyroscopes in an Inertial Motion Unit (IMU) from 42.5 kms of vehicle driving where RSF were present. In particular, two types of RSF were investigated: a rumble strip and a speed bumper. Two different approaches were used to classify the RSF: a feature-based approach derived from the research literature and a novel approach based on dynamic time warping (DTW). The results show that the approach based on Dynamic Time Warping (DTW) provides a very high level of identification accuracy. To the best of the authors' knowledge, this is the first attempt in the literature for the identification of specific road irregularities based on a large experimental dataset.