Search Results

The goal of this study was to use naturalistic driving data to characterize the motion of vehicles making right turns at signalized intersections. Right-turn maneuvers from 13 intersections were extracted from the Second Strategic Highway Research Program (SHRP2) database and categorized based on whether or not the vehicle came to a stop prior to making its turn. Out of the vehicles that did stop, those that were the first and second in line at the intersection were isolated. This resulted in 186 stopped first-in-line turns, 91 stopped second-in-line turns, and 353 no stop turns. Independent variables regarding the maneuver, including driver’s sex and age, vehicle type, speed, and longitudinal and lateral acceleration were extracted. The on-board video was reviewed to categorize the road as dry/wet and if it was day/night. Aerial photographs of the intersections were obtained, and the inner radius of the curve was measured using the curb as a reference.

Passenger vehicle bumpers are designed to reduce collision damage. If colliding bumpers are not vertically aligned, their effectiveness is reduced and the resulting damage increases. Two bumpers of similar static design heights may become misaligned due to bumper dive caused by one or both vehicles pitching forward due to braking. Previous researchers have quantified bumper dive and how it changed with passenger vehicle designs. Currently there are limited data available to quantify the mean, variance, and distribution of bumper dive for modern ABS-equipped vehicles. We conducted maximum braking tests using 3 late-model minivans/CUVs (crossover utility vehicles) and 9 late-model sedans on contiguous dry asphalt and concrete road surfaces. Between 16 and 23 tests were conducted for each vehicle and all tests were conducted from an initial speed of about 65 km/h (40 mph). A laser distance sensor mounted to the front bumpers measured bumper height throughout each test.

Bicycle computers record and store global position data that can be useful for forensic investigations. The goal of this study was to estimate the absolute error of the latitude and longitude positions recorded by a common bicycle computer over a wide range of riding conditions. We installed three Garmin Edge 530 computers on the handlebars of a bicycle and acquired 9 hours of static data and 96 hours (2214 km) of dynamic data using three different navigation modes (GPS, GPS+GLONASS, and GPS+Galileo satellite systems) and two geographic locations (Vancouver, BC, Canada and Orange County, CA, USA). We used the principle of error propagation to calculate the absolute error of this device from the relative errors between the three pairs of computers. During the static tests, we found 16 m to 108 m of drift during the first 4 min and 1.4 m to 5.0 m of drift during a subsequent 8 min period. During the dynamic tests, we found a 95th percentile absolute error for this device of ±8.04 m.

Video evidence in collision reconstruction has become a common foundation for vehicle position and speed analyses. The goal of this study was to explore how the uncertainty of these position/speed analyses is affected by various camera-, scene-, and vehicle-related properties. To achieve this goal, we quantified how the size and aspect ratio of pixels in the pixel grid change as a result of correcting for lens distortion and projecting the pixel grid onto a real-world surface captured by the image. Relying on both general and case-specific examples, we used Monte Carlo analyses to explore how uncertainty can be calculated and how it varies for different measurements and different camera-, scene-, and vehicle-related properties.

The accuracy of collision severity data recorded by event data recorders (EDRs) has been previously measured primarily using barrier impact data from compliance tests and experimental low-speed impacts. There has been less study of the accuracy of EDR-based collision severity data in real-world, vehicle-to-vehicle collisions. Here we used 189 real-world front-into-rear collisions from the Crash Investigating Sampling System (CISS) database where the EDR from both vehicles recorded a severity to examine the accuracy of the EDR-reported speed changes. We calculated relative error between the EDR-reported speed change of each vehicle and a speed change predicted for that same vehicle using the EDR-reported speed change of the other vehicle and conservation of momentum. We also examined the effect of vehicle-type, mass ratio, and pre-impact braking on the relative error in the speed changes.

Anti-lock brake systems (ABS) produce high levels of vehicle deceleration under emergency braking conditions by modulating tire slip. Currently there are limited data available to quantify the mean, variance, and distribution of vehicle deceleration levels for modern ABS-equipped vehicles. We conducted braking tests using twenty (20) late-model vehicles on contiguous dry asphalt and concrete road surfaces. All vehicles were equipped with a 5th wheel sampled at 200 Hz, from which vehicle speed and deceleration as a function of time were calculated. Eighteen (18) tests were conducted for each vehicle and all tests were conducted from a targeted initial speed of 65 km/h (40 mph). Overall, we found that late-model ABS-equipped vehicles can decelerate at average levels that vary from about 0.871g to 1.081g across both surfaces, and that deceleration levels were on average about 0.042g higher on asphalt than on concrete.

Vehicle rollovers are complex events that can be difficult to reconstruct. The goal of this study was to explore whether different vehicle trip models could identify when during the trip phase a vehicle possesses the dynamic conditions needed to rollover. We used three sport utility vehicles (SUVs) with either absent or disabled electronic stability control to conduct six tests involving a steer-induced control loss on a large flat concrete surface. Vehicle kinematics were measured using a GPS speed sensor, tri-axial accelerometers, tri-axial angular rate sensors, and both drone- and land-based video cameras. Four vehicle trip metrics were derived and evaluated using the vehicle dynamics between steer onset and the end of the trip phase. During three tests, one or more of the vehicle’s tires lifted off the ground but the vehicle did not roll. In the other three tests, the vehicle rolled.

The goal of this study was to use naturalistic driving data to characterize the longitudinal and lateral accelerations of vehicles making a left turn from a stop at signalized intersections. Left turn maneuvers at 15 intersections were extracted from the Second Strategic Highway Research Program (SHRP2) database. A subset of 420 traversals for lead vehicles that were initially stopped and negotiated their left turns unimpeded by oncoming traffic was used for the analysis. For each traversal, we extracted information regarding the driver’s sex and age, the vehicle type, the vehicle’s longitudinal and lateral acceleration, and on-board forward-facing video. From the video, we further extracted information about whether the road was dry/wet and if it was day/night, and from aerial photographs of the intersections we extracted the radius of each left turn path through the intersection.

Rollover crashes are complex events that generate motions in all six degrees of freedom (6DOF). Directly quantifying the angular rotations from video can be difficult and vehicle orientation as a function of time is often not reported for staged rollover crashes. Our goal was to evaluate the ability of using a match-moving technique and consumer-grade video cameras to quantify the roll, pitch and yaw angles and angular velocities of a rollover crash. We staged a steer-induced rollover of an SUV at 106 km/h. The vehicle was fitted with tri-axial accelerometers and angular rate sensors, and five consumer-grade video cameras (2 on tripods, 2 on drones, 1 handheld, ~30 fps) captured the event. Roll, pitch and yaw angles were determined from the video using specialized software.

Data downloaded from event data recorders (EDRs) integrated into the airbag systems of passenger vehicles can be key evidence for collision investigators. Often the EDR data includes information about the severity of the collision in terms of the longitudinal and lateral speed changes experienced by the vehicle. Previous studies have shown that for collisions with small lateral speed changes, the accuracy of the reported longitudinal speed change varies with manufacturer and magnitude. The goal of this study was to quantify the accuracy of EDR-reported speed changes in high-speed angled collisions with larger lateral speed change components. Data from 25 crash tests conducted for the National Highway Traffic Safety Administration’s (NHTSA’s) Oblique Offset Frontal Impact Research and Development Program were used in this study.

Modern vehicles record dynamic data from a number of on-board sensors for events that could precede a crash. These data can be used to reconstruct the behavior of a vehicle, although the accuracy of these reconstructions has not yet been quantified. Here, we evaluated various methods of reconstructing the vehicle kinematics of a 2017 and a 2018 Toyota Corolla based on Vehicle Control History (VCH) data from overlapping events generated by the pre-collision system (PCS), sudden braking (SB) and anti-lock brake (ABS) activation. The vehicles were driven towards a stationary target at 32-64 km/h (20-40 mph) and then after the pre-collision alarm sounded the vehicle was steered sharply right or left and braked rapidly to rest. VCH data for PCS event were recorded at 2 Hz and for the sudden braking and ABS activation events at 6.7 Hz.

Newer Toyota vehicles store information about more than 50 parameters for 5 s before and after non-collision events in the Vehicle Control History (VCH) records. The goals of this study were to assess the accuracy of VCH data acquired during Autonomous Emergency Braking (AEB) events and to investigate the effects of speed, acceleration, and system settings on AEB performance. A 2017 Toyota Corolla with Safety Sense P Pre-Collision System (PCS) was driven in a straight line towards a car-like target at different combinations of four speeds (20, 25, 30, and 40 km/h; or 12, 15, 19, and 25 mph) and three accelerator pedal positions (constant 30%, 40%, and 50% accelerator opening ratios) until the AEB system activated. The vehicle speed, vehicle acceleration, radar target closing speed, and radar target distance recorded in the VCH were compared to a reference 5th wheel. We found that errors in the VCH distance, speed, and acceleration data varied with the test conditions.

The Pre-Collision System (PCS) in Toyota’s Safety Sense package includes an autonomous emergency braking feature that can stop or slow a vehicle independent of driver input if there is an impending collision. The goals of this study were to determine how hazard characteristics, specifically radar reflector size and degree of target edge contrast, affect the response of the PCS, as well as to scrutinize tests wherein the PCS failed to stop the vehicle before impact. We conducted 80 tests with a 2017 Toyota Corolla driven towards a car-like target in a straight line and under constant accelerator pedal position, reaching about 30 km/h at the PCS alarm. Vehicle speed and distance to target at the alarm flag (ALM) and at times corresponding to three other system flags (PBA, FPB, and PB) were read from the Vehicle Control History records. Time to impact (TTI) at each flag was calculated and the distance between the stopped vehicle and the target was measured for each test.

The repeatability and accuracy of front and rear speed changes reported by Toyota’s Airbag Control Modules (ACMs) have been previously characterized for low-severity collisions simulated on a linear sled. The goals of the present study are (i) to determine the accuracy and repeatability of Toyota ACMs in mid-severity crashes, and (ii) to validate the assumption that ACMs function similarly for idealized sled pulses and full-scale vehicle-to-barrier and vehicle-to-vehicle crashes. We exposed three Toyota Corollas to a series of full-scale aligned frontal and rear-end crash tests with speed changes (ΔV) of 4 to 12 km/h. We then characterized the response of another 16 isolated Toyota ACMs from three vehicle models (Corolla, Prius and Camry) and 3 generations (Gen 1, 2 and 3) using idealized sled pulses and replicated vehicle-to-vehicle and vehicle-to-barrier pulses in both frontal and rear-end crashes (ΔV = 9 to 17 km/h).

The ability to accurately calculate a snowmobile’s speed based on measured track marks in the snow is important when assessing a snowmobile accident. The characteristics and length of visible snowmobile track marks were documented for 41 locked-track braking tests and 38 rolldown tests using four modern snowmobiles on a groomed/packed snow surface. The documented track mark lengths were used to quantify the uncertainty associated with using track mark length to estimate initial speed. Regression models were developed for both data sets. The regression model of the locked-track tests revealed that using an average deceleration of 0.36g over the length of the locked track mark provides a good estimate of the best-fit line through the data, with the upper and lower 95th percentile prediction interval bounds best represented by using deceleration rates of 0.23g and 0.52g respectively.

The accuracy of the speed change reported by Generation 1 Toyota Corolla Event Data Recorders (EDR) in low-speed front and rear-end collisions has previously been studied. It was found that the EDRs underestimated speed change in frontal collisions and overestimated speed change in rear-end collisions. The source of the uncertainty was modeled using a threshold acceleration and bias model. This study compares the response of Generation 1, 2 and 3 Toyota EDRs from Toyota Corolla, Camry and Prius models. 19 Toyota airbag control modules (ACMs) were mounted on a linear sled. The ACMs underwent a series of frontal and rear-end haversine crash pulses of varying severity, duration and peak acceleration. The accuracy and trigger thresholds of the different models and generations of EDRs were compared. There were different accuracy trends found between the early Generation 1 and the more modern Generation 2 and 3 EDRs.

There are limited scientific data available on the acceleration and braking performance of modern snowmobiles. In this study we investigated the acceleration and deceleration characteristics of four modern snowmobiles of varying engine size (500 to 1000 cc) and style (2-stroke and 4-stroke) on groomed/packed snow conditions. The acceleration tests were performed at quarter, half and full throttle. The deceleration tests were performed using full braking with locked tracks and rolldown with power both on and off. Target test speeds ranged from 20 to 60 km/h. Snow condition parameters were measured throughout the tests. The results of the acceleration tests showed that at higher speeds, higher horsepower rating generally corresponds to higher acceleration rates, with a maximum observed average acceleration of 0.70g.

In low-speed collisions, motor vehicles can lose a significant fraction of their initial kinetic energy without plastic deformation or damping elements in their bumper assemblies. Five vehicles were subjected to multiple, non-damaging barrier and vehicle-to-vehicle impacts. Position, velocity, acceleration and force data were recorded for all collisions. Modeling vehicles as non-rigid two degree of freedom systems accurately predicted velocity and restitution responses for five vehicles in barrier and vehicle-to-vehicle impacts.

Crash data recorded by the restraint control module (RCM) installed in newer Ford passenger vehicles have recently become available to investigators. To quantify the accuracy of the crash data in low-speed collisions, two RCM-equipped vehicles were exposed to 84 aligned frontal barrier collisions with speed changes up to 13.5 km/h. The accuracy of the speed change reported by the RCM ranged from an underestimate of 1.8 km/h to an overestimate of 0.3 km/h. The error varied with speed change. The RCMs were mounted on a linear sled to investigate their sensitivity to specific collision pulse parameters. For both RCMs, the first eight acceleration data points were duplicated at the end of the data and the record of the crash pulse was often incomplete. Based on the results of this study, crash investigators need to carefully interpret the RCM-reported acceleration and speed change data before using it to reconstruct low-speed collisions involving Ford vehicles.

Adrian's visibility model is a useful tool for assessing the visibility of an object at night. However, it was developed under laboratory conditions. Thus, it is necessary to determine the visibility levels which are required for detection under nighttime driving conditions. Experimental data from Olson et al were applied to the Adrian visibility model to determine visibility levels at target detection for alerted drivers. The data has been modified to account for experimental delay in the recorded detection points and a correction has been applied to assess driver expectation. Driver age, headlight beam pattern, and target reflectivity were all found to have a significant effect on visibility level at target detection. For alerted drivers, 50th-percentile threshold visibility levels between 1 and 23 were calculated. For unalerted drivers, 50th-percentile threshold visibility levels between 13 and 210 were calculated.