Oscar Sánchez Siordia

Driving risk classification based on experts evaluation

A novel multidisciplinary system for the automatic driving risk level classification is presented. The data considered involves the three basic traffic safety elements (driver, road, and vehicle), as well as knowledge from traffic experts. The driving experiments were conducted in a truck cabin simulator handled by a professional driver, considering the most common real-world enviroments. Each traffic expert evaluate the driving risk on a 0 to 100 visual analogue scale. The driver, road and vehicle information was used to train five different data mining algorithms in order to predict the driving risk level. The benefits of the completeness of the data considered in our system are presented and discussed.

Accident reproduction system for the identification of human factors involved on traffic accidents

In this paper, a novel accident reproduction system for the identification of the main human factors involved on traffic accidents is presented. The system is based on a wireless in-vehicle Electronic Data Recorder that could be easily installed in any vehicles cabin for the monitoring of the three basic elements of traffic safety: driver, road and vehicle. The system has been tested in a highly realistic truck simulator with a group of professional drivers. The data, collected with the system at the moments before traffic accidents, were used to generate a novel database that was carefully analyzed by a group of traffic safety experts. The validation process shows the reliability of the developed system as a tool for the identification of the main causes of the monitored traffic accidents.

Section-wise similarities for clustering and outlier detection of subjective sequential data

In this paper, a novelty methodology for the representation and similarity measurement of sequential data is presented. First, a linear segmentation algorithm based on feature points is proposed. Then, two similarity measures are defined from the differences between the behavior and the mean level of the sequential data. These similarities are calculated for clustering and outlier detection of subjective sequential data generated through the evaluation of the driving risk obtained from a group of traffic safety experts. Finally, a novel dissimilarity measure for outlier detection of paired sequential data is proposed. The results of the experiments show that both similarities contain complementary and relevant information about the dataset. The methodology results useful to find patterns on subjective data related with the behavior and the level of the data.

Combining traffic safety knowledge for driving risk detection

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.

Subjective Traffic Safety Experts' Knowledge for Driving-Risk Definition

This paper presents a novelty system for the detection of driving-risk situations based on the knowledge acquired from traffic safety experts. A complete methodology to generate a driving-risk reference signal has been developed. A set of driving sessions was executed in a very realistic truck simulator, where several measures and visual information from the vehicle, the driver, and the road were collected. Two kinds of experiments were designed, i.e., controlled driving sessions (where several risk situations were induced) and natural driving sessions (where no risk situations were induced and a natural driving behavior was expected). A group of traffic safety experts from the Royal Automobile Club of Spain was consulted to evaluate the driving risk in each simulated session. The information acquired from the traffic safety experts was used to develop a methodology to combine their evaluations. The risks detected with the proposed methodology were analyzed to determine the most common human factors related with the generation of driving-risk situations.

Wireless In-Vehicle Compliant Black Box: For Accident Analysis

In this article, a novel accident analysis system is proposed to determine the main human factors involved in traffic accidents. Several tests of the system were carried out in a highly realistic truck simulator with a group of professional drivers. The data, collected with the system at the moments before traffic accidents, were used to generate a novel database that was carefully analyzed by a group of traffic safety experts. The validation process shows the reliability of the developed system as a tool for the identification of the main cause of traffic accidents.

Combining experts knowledge for driving risks recognition

In this paper, a methodology to recognize driving risk situations as the solution of a combination of information problem is presented. A collection of simulated sessions in a highly realistic truck simulator were designed and executed. Several internal truck magnitudes and visual information from the driver and the road were collected in each session. Two traffic safety experts were asked to evaluate the driving risk of the exercises using a simulation reproduction tool (developed for this purpose) and a Visual Analog Scale (VAS). These evaluations were used to define four different, and complementary, models for driving risk recognition. A method to calculate these models by the maximization of a similarity measure between expert evaluations is presented. Finally, a third traffic safety expert was consulted for validation purposes. Results show that the proposed models are useful and able to recognize abnormal drivers behavior. Good generalization results were obtained when the parameters learned for each risk definition were validated in additional simulated sessions.

Section-Wise similarities for classification of subjective-data on time series

The aim of this paper is to present a novelty methodology to develop similarity measures for classification of time series. First, a linear segmentation algorithm to obtain a section-wise representation of the series is presented. Then, two similarity measures are defined from the differences between the behavior of the series and the level of the series, respectively. The method is applied to subjective-data on time series generated through the evaluations of the driving risk from a group of traffic safety experts. These series are classified using the proposed similarities as kernels for the training of a Support Vector Machine. The results are compared with other classifiers using our similarities, their linear combination and the raw data. The proposed methodology has been successfully evaluated on several databases.

Automatic driving risk detection based on hands activity

In this paper a novelty methodology to measure driving risk based on hands activity is presented. The proposed algorithm has been developed and tested on several driving sessions executed on a highly realistic truck simulator. The hands positions are used to feed a risk buffer that is in charge of penalizing wrong hands activities and praising good hands activities to generate a measure of the driving risk. In order to select the parameters of the proposed system, a genetic algorithm (GA) and a ground truth acquired from a group of traffic safety experts were considered. The results of the proposed methodology on several driving sessions showed its effectiveness to automatically detect risky situations related to bad drivers hands behavior. The present system will be integrated in a global alarm system that will be included in an intelligent truck cabin for road transportation.

Subjective data arrangement using clustering techniques for training expert systems

Abstract The evaluation of subjective data is a very demanding task. The classification of the information gathered from human evaluators and the possible high noise levels introduced are ones of the most difficult issues to deal with. This situation leads to adopt individuals who can be considered as experts in the specific application domain. Thus, the development of Expert Systems (ES) that consider the opinion of these individuals have been appeared to mitigate the problem. In this work an original methodology for the selection of subjective sequential data for the training of ES is presented. The system is based on the arrangement of knowledge acquired from a group of human experts. An original similarity measure between the subjective evaluations is proposed. Homogeneous groups of experts are produced using this similarity through a clustering algorithm. The methodology was applied to a practical case of the Intelligent Transportation Systems (ITS) domain for the training of ES for driving risk prediction. The results confirm the relevance of selecting homogeneous information (grouping similar opinions) when generating a ground truth (a reliable signal) for the training of ES. Further, the results show the need of considering subjective sequential data when working with phenomena where a set of rules could not be easily learned from human experts, such as risk assessment.