An evaluation study of driver profiling fuzzy algorithms using smartphones

Abstract

Profiling driving behavior has become a relevant aspect in fleet management, automotive insurance and eco-driving. Detecting inefficient or aggressive drivers can help reducing fleet degradation, insurance policy cost and fuel consumption. In this paper, we present a Fuzzy-Logic based driver scoring mechanism that uses smartphone sensing data, including accelerometers and GPS. In order to evaluate the proposed mechanism, we have collected traces from a testbed consisting in 20 vehicles equipped with an Android sensing application we have developed to this end. The results show that the proposed sensing variables using smartphones can be merged to provide each driver with a single score.

An inference engine for smartphones to preprocess data and detect stationary and transportation modes

“…Using Accelerometer and GSM without any report for motorized mode detection Hong et al (2014) Driver behavior Recognition They are activated by users Connection with OBD Dai et al (2010) Manual activation and installation in each trip Wahlstrom et al (2015) Johnson and Trivedi (2011) Castignani et al (2013) detect and analyze stationary states were drawbacks of their research. Gong et al (2012) used a GPS device and geographical information from ArcGIS to determine the modes based on recorded locations and speeds.…”

SenseFleet: A smartphone-based driver profiling platform

“…In the following we provide a brief description of the SenseFleet algorithm. For more a more detailed description the reader is referred to our previous works [3], [4].…”

Driver Behaviour Profiling Using Dynamic Bayesian Network

“…A comparative experimentation of the DriveSafe [9] and the adaptive fuzzy model [10] saw the fuzzy model significantly reduce the number of false detections with considerable increase in real detections. Castignani, Frank and Engel [11] had early carried out an evaluation study of driver profiling fuzzy algorithms using GPS, accelerometer, magnetometer and gravity smartphone sensor. Their [11] proposed model puts all event types at the same scoring priority level merged in a global event counter.…”

“…Castignani, Frank and Engel [11] had early carried out an evaluation study of driver profiling fuzzy algorithms using GPS, accelerometer, magnetometer and gravity smartphone sensor. Their [11] proposed model puts all event types at the same scoring priority level merged in a global event counter. Sensor data from the smartphone is first filtered with events detected for the different metrics, followed by input data fuzzification with fuzzy rules applied in a Fuzzy Inference Engine [11].…”