Combining traffic safety knowledge for driving risk detection

Abstract

In this paper a novelty method to combine knowledge of traffic safety experts, in order to detect driving risk situations, is presented. A set of driving sessions were executed in a very realistic truck simulator where several magnitudes and visual information from the vehicle, driver and road were collected. Two kind of experiments were designed: controlled driving sessions (where several risky situations were induced), and natural driving sessions (where a natural driving behavior was expected). A group of traffic safety experts were consulted to evaluate the driving risk in each session. The information acquired from the traffic safety experts was used to develop a methodology to combine the information and to define a set of driving risk models. The developed system detected most of the induced risk situations besides of several non-induced risk situations. The methodology presented in this paper can be used to obtain a driving risk ground truth in order to compare and evaluate risk detection algorithms and to analyze the influence of vehicle, driver and road variables on the driving risk.

Optimal experts' knowledge selection for intelligent driving risk detection systems

“…In this case, an algorithm to represent experts' VAS evaluations into its simplest form, to easily compare and combine their information, was applied. This algorithm, called Trend Segmentation Algorithm (T SA), has been successfully used for the linear representation of subjective data (see, for instance, [9], [16] and [17]). In a nutshell, given a set of feature points of the VAS evaluations where the trend of the data presents a deviation from a straight course, the TSA algorithm looks for a set of points where a piecewise linear model can be properly fitted (see [16], for a complete description of TSA).…”

“…, 31}) combinations, were considered. For each combination, a driving risk ground truth was generated following [9]. Given The Level Risk (LR) is set to 1 if the mean levels of most of the TAS evaluations in the considered section are higher than a threshold (set to 50 in our case).…”

“…Expert panels have been and are being used extensively in various research programs related to the development of the Highway Safety Manual and deserve the careful scrutiny that other methods receive [8]. In [9], a method for the generation of a ground truth for driving risk measurement based on the combination of knowledge acquired from traffic safety experts is presented. The combination is achieved using a computer-facilitated collaborative work approach for the integration of experts' opinions.…”