Pre-crash Data Research Articles

Understanding the role of driver behaviors and performance in safety-critical events: Application of machine learning

Human factors contribute in some way to almost 93% of road crashes. Because the unavailability of good pre-crash data, the contribution of human factors to safety-critical events (SCEs) and the prediction of crashes using real-world data is lightly researched. This study provides predictive accuracy by harnessing unique real-world naturalistic driving study (NDS) data, which includes dynamic pre-crash information about driving behavior and performance. After cleaning and preprocessing, a final subsample (N = 9,237) was used and split into training and test samples. For consistent comparison of variables’ importance in statistical and machine learning (ML) models, the dominance analysis uncovered the most important predictors used by the ordered Probit model. Next, three non-parametric supervised ML methods, because promising prediction performance and cost-effectiveness, including Naïve Bayes, K-Nearest Neighbors, and Gradient Boosting Tree (GBT) were used. The overall out-of-sample prediction accuracy for the ordered Probit model was 85.75% which was lower than all three ML methods. The GBT showed the highest (91.23%) out-of-sample prediction accuracy. The availability of pre-crash naturalistic data helps significantly improve the prediction accuracy of SCEs as cumulative importance for all available human factors in the GBT classifier was 94%. For practical applications, refer to the article.

Inferring safety critical events from vehicle kinematics in naturalistic driving environment: Application of deep learning Algorithms

Advances in sensing technology has enabled the collection of countless terabytes of second-by-second kinematics data. Such data provides opportunities for real-time monitoring of driving behavior and identification of safety critical events (SCEs) including crashes and near crashes. The concept of volatility is relevant in this context, which identifies instability and erratic variations in driving behavior prior to involvement in SCEs. This study utilized vehicle kinematics from a large-scale naturalistic driving data to develop a deep learning approach based on 1D convolutional neural networks (CNN) for inferring SCEs. The data are unique in the sense that such accurate pre-crash data at high fidelity are not available in traditional crash repositories. This study contributes to the literature by providing a first attempt at predicting responses to SCEs by developing deep learning-based CNN architectures using novel driving volatility based kinematic thresholds for a sample of 9553 events. The key contribution lies in developing a volatility-based CNN input layout that is acceptable to CNN schemes and represents the motion kinematics such as speed, acceleration and volatility measures. Several 1D-CNN architectures were developed using layers, numbers of convolutions, layer patterns, and kernels. Shallow and deep architectures were tested, revealing higher accuracy of shallow architectures in detecting SCEs. The optimal number of epochs were identified using an early stopping method while the CNN performance was improved by increasing the number of epochs. The ensemble CNN had the highest predictive accuracy of 95.6% for detection of crashes and near crashes, which was 2.5% higher than the optimal CNN using 20% hold out test data. The ensemble CNN also outperformed classical machine learning models and model performance reported in past studies on detection of SCEs. These results have implications for identification of safety hotspots and providing real-time alerts and warnings in connected and highly automated vehicle environment including society of automotive engineers levels 3–5.

Construction of Collision-Type Prediction Models Based on Pre-crash Data for Advanced Driver Assistance Systems

ABSTRACT: Collision-type prediction models based on pre-crash information are important because there is a relationship between collision type, avoidance operations, and occupant injuries. Thus, they can be applied to autonomous driving systems (ADS) or advanced driver assistance systems (ADAS) to prevent serious accidents or minimize damage during collisions. In this study, we investigated the application of collision-type prediction models based on several machine learning methods and compared their performance to determine the best model based on their f1 scores. The results revealed that the light gradient boosting machine (LGBM) model had a high f1 score that exceeded 0.92, which implied that it could potentially be used for ADS and ADAS applications. Furthermore, a brief analysis was performed on the ranking of various factors, which provided useful insight into the significance of several pre-crash factors and their distributions.

Investigating the Effect of Social and Cultural Factors on Drivers in Malaysia: A Naturalistic Driving Study.

Road accidents are increasing every year in Malaysia, and it is always challenging to collect reliable pre-crash data in the transportation community. Existing studies relied on simulators, police crash reports, questionnaires, and surveys to study Malaysia’s drivers’ behavior. Researchers previously criticized such methods for being biased and unreliable. To fill in the literature gap, this study presents the first naturalistic driving study in Malaysia. Thirty drivers were recruited to drive an instrumented vehicle for 750 km while collecting continuous driving data. The data acquisition system consists of various sensors such as OBDII, lidar, ultrasonic sensors, IMU, and GPS. Irrelevant data were filtered, and experts helped identify safety criteria regarding multiple driving metrics such as maximum acceptable speed limits, safe accelerations, safe decelerations, acceptable distances to vehicles ahead, and safe steering behavior. These thresholds were used to investigate the influence of social and cultural factors on driving in Malaysia. The findings show statistically significant differences between drivers based on gender, age, and cultural background. There are also significant differences in the results for those who drove on weekends rather than weekdays. The study presents several recommendations to various public and governmental sectors to help prevent future accidents and improve traffic safety.

Evaluation of the efficacy of an intersection conflict warning system at two-way stop-controlled rural intersections: difference-in-differences and triple-difference analytical approaches

ObjectiveIntersection conflict warning systems (ICWSs) have been implemented at high-risk two-way stop-controlled intersections to prevent right-angle crashes and associated injuries. This study involved investigation of the impacts of ICWSs on...

Driving impairments and duration of distractions: Assessing crash risk by harnessing microscopic naturalistic driving data

Distracted and impaired driving is a key contributing factor in crashes, leading to about 35% of all transportation-related deaths in recent years. Along these lines, cognitive issues like inattentiveness can further increase the chances of crash involvement. Despite its prevalence and importance, little is known about how the duration of these distractions is associated with critical events, such as crashes or near-crashes. With new sensors and increasing computational resources, it is possible to monitor drivers, vehicle performance, and roadway features to extract useful information, e.g., eyes off the road, indicating distraction and inattention. Using high-resolution microscopic SHRP2 naturalistic driving data, this study conducts in-depth analysis of both impairments and distractions. The data has more than 2 million seconds of observations in 7394 baselines (no event), 1228 near-crashes, and 617 crashes. The event data was processed and linked with driver behavior and roadway factors. The intervals of distracted driving during the period of observation (15 seconds) were extracted; next, rigorous fixed and random parameter logistic regression models of crash/near-crash risk were estimated. The results reveal that alcohol and drug impairment is associated with a substantial increase in crash/near-crash event involvement of 34%, and the highest correlations with crash risk include duration of distraction through dialing on a cellphone, texting while driving, and reaching for an object. Using detailed pre-crash data from instrumented vehicles, the study contributes by quantifying crash risk vis-à-vis detailed driving impairment and information on secondary task involvement, and discusses the implications of the results.

Event Data Recorder Performance during High Speed Yaw Testing Subsequent to a Simulated Tire Tread Separation Event

<div class="section abstract"><div class="htmlview paragraph">This paper presents event data from the Sensing and Diagnostic Module (SDM) of a 2004 Chevrolet Malibu during high speed yaw testing. Yaw tests were performed using tires that were intact and tires that had the tread removed. The tires that had the tread removed were placed at various wheel positions on the vehicle (e.g. leading side - front, leading side -rear, trailing side - rear). This testing simulates the loss of control phase subsequent to a tread separation. Speeds up to 117 km/h (72.9 mph) were achieved. A simple electro-mechanical device was incorporated to the dynamic testing to simulate a low-severity non-deployment event that triggered the recording of pre-crash data by the SDM. The SDM data from the tests was imaged and compared to reference data from vehicle-mounted instrumentation recording wheel speed, steering angle, measured vehicle sideslip angle and GPS calculated over the ground speed. This paper examines the dynamic effect of high sideslip angles and changes to tire rolling radius, as a result of a tread separation, as it pertains to the accuracy of EDR reported vehicle speed.</div></div>

Models of Driver Acceleration Behavior Prior to Real-World Intersection Crashes

Drivers involved in intersection collisions are at high risk of serious or fatal injury. Intersection advanced driver assistance systems (I-ADAS) are emerging active safety systems designed to help drivers safely traverse intersections. The effectiveness of I-ADAS is expected to be greatly dependent on pre-crash vehicle acceleration during intersection traversals. The objective of this paper was to develop pre-crash acceleration models for non-turning drivers involved in straight crossing path crashes and left-turning drivers in left turn across path opposite direction and lateral direction crashes. This paper used 348 event data recorder pre-crash records taken from crashes investigated as part of the National Automotive Sampling System/Crashworthiness Data System. The acceleration models generated from this pre-crash data were evaluated using a leave-one-out cross-validation procedure. Previously developed non-crash models from the literature were compared with the pre-crash models. Our hypothesis was that drivers involved in crashes would accelerate more aggressively than the “typical” driving population. This result suggests that drivers in pre-crash scenarios tend to accelerate more aggressively than drivers in normal scenarios (p<0.001). This has important implications for the design of I-ADAS. Specifically, higher acceleration results in less available time for I-ADAS to detect and respond to an imminent collision.

Vbox와 PC-Crash를 활용한 EDR 기록정보의 신뢰성 평가

The EDR(Event Data Recorder) is a part of the ACU(Airbag Control Unit) functions mounted on a vehicle. EDR data have pre-crash data and post-crash data. Pre-crash data are recorded within 5 sec from time zero(AE) with 0.5 sec resolution, and reveal vehicle speed, engine rotation speed, throttle opening, brake pedal operation, acceleration pedal position and steering angle, etc. Using this EDR data, the investigation of a traffic accident can become more objective and scientific. Crash tests of three vehicles equipped with EDR function had been performed successfully. Evaluation of EDR data reliability had also been performed using Vbox and PC-Crash’s sequence table function. Based on the results, we could confirm EDR data’s reliability and availability for Traffic Accident Analysis by the series of this process.

Context-aware system for pre-triggering irreversible vehicle safety actuators

New vehicle safety systems have led to a steady improvement of road safety and a reduction in the risk of suffering a major injury in vehicle accidents. A huge leap forward in the development of new vehicle safety systems are actuators that have to be activated irreversibly shortly before a collision in order to mitigate accident consequences. The triggering decision has to be based on measurements of exteroceptive sensors currently used in driver assistance systems. This paper focuses on developing a novel context-aware system designed to detect potential collisions and to trigger safety actuators even before an accident occurs. In this context, the analysis examines the information that can be collected from exteroceptive sensors (pre-crash data) to predict a certain collision and its severity to decide whether a triggering is entitled or not. A five-layer context-aware architecture is presented, that is able to collect contextual information about the vehicle environment and the actual driving state using different sensors, to perform reasoning about potential collisions, and to trigger safety functions upon that information. Accident analysis is used in a data model to represent uncertain knowledge and to perform reasoning. A simulation concept based on real accident data is introduced to evaluate the presented system concept.

Changes in baseline concussion assessment scores following a school bus crash

ABSTRACTObjective: The objective of this article is to present concussion assessment data for 30 male athletes prior to and after being involved in a large school bus crash. The athletes on the bus, all male and aged 14–18 years, were participants in their school's concussion management program that included baseline and postinjury testing using Immediate Post-Concussion Assessment and Cognitive Testing (ImPACT).Methods: This case study described changes in concussion assessment scores for 30 male athletes following a primarily frontal school bus crash. Data from the school's concussion management program, including baseline test data and postinjury assessment data, were reviewed. Athletes who required multiple postinjury assessments by the program were identified as having had significant cognitive changes as a result of the bus crash.Results: Twenty-nine of 30 athletes were injured. One had lumbar compression fractures; others had various lacerations, abrasions, contusions, sprains, and nasal fractures. ImPACT data (postcrash) were available for all 30 athletes and 28 had available precrash baseline data. A total of 16 athletes (53.3%) had significant cognitive changes indicated by changes in their concussion assessment scores, some of which took months to improve.Conclusion: This case study highlights a unique opportunity to evaluate concussion assessment data from 30 male athletes involved in a high-speed school bus crash. Further, these data provide additional insight into assessing the effectiveness of current school bus occupant protection systems.

Foreign direct investment in a small open economy

Think of a small open economy interestingly positioned between the trade blocs of the NAFTA and the EU, with FDI in recent years resembling the pattern before the economic crash, making a pre-crash data set useful for exploring potential long-term trends. In this research, investment is explained by geographic location and country size, using a gravity model to account for the country's exceptional remoteness and sparseness. A unique extension of the gravity model applies the inverse hyperbolic sine (IHS) function. The IHS functional form is estimated together with fixed difference between investment sectors and trade blocs being estimated simultaneously, an analysis that is rarely possible. Results indicate that under these conditions, investment appears to be more driven by wealth than market size effects.

Characterization of Lane Departure Crashes Using Event Data Recorders Extracted from Real-World Collisions

Lane Departure Warning (LDW) is a production active safety system that can warn drivers of an unintended departure. Critical in the design of LDW and other departure countermeasures is understanding pre-crash driver behavior in crashes. The objective of this study was to gain insight into pre-crash driver behavior in departure crashes using Event Data Recorders (EDRs). EDRs are units equipped on many passenger vehicles that are able to store vehicle data, including pre-crash data in many cases. This study used 256 EDRs that were downloaded from GM vehicles involved in real-world lane departure collisions. The crashes were investigated as part of the NHTSA's NASS/CDS database years 2000 to 2011. Nearly half of drivers (47%) made little or no change to their vehicle speed prior to the collision and slightly fewer decreased their speed (43%). Drivers who did not change speed were older (median age 41) compared to those who decreased speed (median age 27). Drivers in lane departure crashes were traveling above the posted speed limit in 65% of cases. Almost half (49%) of drivers did not apply the brakes during the pre-crash record, suggesting that drivers may be steering more often than braking to avoid crashes. The newest advanced EDRs record steering wheel angle in addition to the data collected by older EDRs most common in the current fleet. We also present a case study using advanced EDR data to simulate the vehicle's trajectory. This method could be invaluable in benefits estimates of safety systems. Language: en

What can the drivers’ own description from combined sources provide in an analysis of driver distraction and low vigilance in accident situations?

Accident data play an important role in vehicle safety development. Accident data sources are generally limited in terms of how much information is provided on driver states and behaviour prior to an accident. However, the precise limitations vary between databases, due to differences in analysis focus and data collection procedures between organisations. If information about a specific accident can be retrieved from more than one data source it should be possible to combine the available information sets to facilitate data from one source to compensate for limitations in the other(s). To investigate the viability of such compensation, this study identified a set of accidents recorded in two different data sources. The first data source investigated was an accident mail survey and the second data source insurance claims documents consisting predominantly of insurance claims completed by the involved road users. An analysis of survey variables was compared to a case analysis including word data derived from the same survey and filed insurance claims documents. For each accident, the added value of having access to more than one source of information was assessed. To limit the scope of this study, three particular topics were investigated: available information on low vigilance (e.g., being drowsy, ill); secondary task distraction (e.g., talking with passengers, mobile phone use); and distraction related to the driving task (e.g., looking for approaching vehicles). Results suggest that for low vigilance and secondary task distraction, a combination of the mail survey and insurance claims documents provide more reliable and detailed pre-crash information than survey variables alone. However, driving related distraction appears to be more difficult to capture. In order to gain a better understanding of the above issues and how frequently they occur in accidents, the data sources and analysis methods suggested here may be combined with other investigation methods such as in-depth accident investigations and pre-crash data recordings.

Analysis of Pre-Crash Data Transferred over the Serial Data Bus and Utilized by the SDM-DS Module

Analysis of Pre-Crash Data Transferred over the Serial Data Bus and Utilized by the SDM-DS Module

Blood Screening in a Southern Nigeria City: A Case Study with SAVAN

Objective: Commercial motorcycle transportation and motorbike riding in Nigeria is prevalent, and road traffic accidents often result. Characteristic of such accidents is massive blood loss, thus exerting extreme pressure on the blood bank for replenishment and screening. The need to galvanize the system to design a blood bank with minimal bureaucracy and easy access led to screening for blood group. A delay in accessing blood for the victims leads to higher mortality. Method: Our approach was to establish a pre-crash blood data for all auto-bike riders who participated in Save Accident Victims Association of Nigeria (SAVAN, an indigenous, nongovernmental organization) training program. Data used were obtained from 1250 auto-bike riders who volunteered at our workshop. Tile grouping method was used for the screening. Results: Blood group O positive (54.3%) was the most common blood group type among the auto-bike riders studied, with A positive following at 20.3 percent, B positive at 18.8 percent, O negative at 3.7 percent, AB positive at 1.3 percent, B negative at 1.1 percent, and A negative at 0.5 percent. It was observed that none of the volunteers grouped AB negative. Conclusion: Blood group of auto-bike riders, pedestrians, passengers, and all potential victims should be documented in their identification card to facilitate blood transfusion during major crisis or disasters where the facilities for typing are not available.