Carol A. C. Flannagan

Identification and validation of a logistic regression model for predicting serious injuries associated with motor vehicle crashes

A multivariate logistic regression model, based upon National Automotive Sampling System Crashworthiness Data System (NASS-CDS) data for calendar years 1999-2008, was developed to predict the probability that a crash-involved vehicle will contain one or more occupants with serious or incapacitating injuries. These vehicles were defined as containing at least one occupant coded with an Injury Severity Score (ISS) of greater than or equal to 15, in planar, non-rollover crash events involving Model Year 2000 and newer cars, light trucks, and vans. The target injury outcome measure was developed by the Centers for Disease Control and Prevention (CDC)-led National Expert Panel on Field Triage in their recent revision of the Field Triage Decision Scheme (American College of Surgeons, 2006). The parameters to be used for crash injury prediction were subsequently specified by the National Expert Panel. Model input parameters included: crash direction (front, left, right, and rear), change in velocity (delta-V), multiple vs. single impacts, belt use, presence of at least one older occupant (≥ 55 years old), presence of at least one female in the vehicle, and vehicle type (car, pickup truck, van, and sport utility). The model was developed using predictor variables that may be readily available, post-crash, from OnStar-like telematics systems. Model sensitivity and specificity were 40% and 98%, respectively, using a probability cutpoint of 0.20. The area under the receiver operator characteristic (ROC) curve for the final model was 0.84. Delta-V (mph), seat belt use and crash direction were the most important predictors of serious injury. Due to the complexity of factors associated with rollover-related injuries, a separate screening algorithm is needed to model injuries associated with this crash mode.

An improved seating accommodation model with application to different user populations

In this paper a new approach to driver seat position modeling is presented. The equations of the Seating Accommodation Model (SAM) separately predict parameters of the distributions of male and female fore/aft seat position in a given vehicle. These distributions are used together to predict specific percentiles of the combined male and female seat position distribution. The effects of the following vehicle parameters are reflected in the prediction of mean seat position: seat height, steering-wheel-to-accelerator pedal distance, seat cushion angle, and transmission type. The mean and standard deviation of driver population stature are included in the prediction for the mean and standard deviation of the seat position distribution, respectively. SAM represents a new, more flexible approach to predicting fore/aft seat position distributions for any driver population in passenger vehicles. Model performance is good, even at percentiles in the tails of the distribution. (A) For the covering abstract of the conference see IRRD 492369.

Fetal outcome in motor-vehicle crashes: effects of crash characteristics and maternal restraint

OBJECTIVE This project was undertaken to improve understanding of factors associated with adverse fetal outcomes of pregnant occupants involved in motor-vehicle crashes. STUDY DESIGN In-depth investigations of crashes involving 57 pregnant occupants were performed. Maternal and fetal injuries, restraint information, measures of external and internal vehicle damage, and details about the crash circumstances were collected. Crash severity was calculated using vehicle crush measurements. Chi-square analysis and logistic regression models were used to determine factors with a significant association with fetal outcome. RESULTS Fetal outcome is most strongly associated with crash severity (P < .001) and maternal injury (P = .002). Proper maternal belt-restraint use (with or without airbag deployment) is associated with acceptable fetal outcome (odds ratio = 4.5, P = .033). Approximately half of fetal losses in motor-vehicle crashes could be prevented if all pregnant women properly wore seat belts. CONCLUSION Higher crash severity, more severe maternal injury, and lack of proper seat belt use are associated with a higher risk of adverse fetal outcome. These results strongly support recommendations that pregnant women use properly positioned seatbelts.

Effects of vehicle interior geometry and anthropometric variables on automobile driving posture

The effects of vehicle package, seat, and anthropometric variables on posture were studied in a laboratory vehicle mockup. Participants (68 men and women) selected their preferred driving postures in 18 combinations of seat height, fore-aft steering wheel position, and seat cushion angle. Two seats differing in stiffness and seat back contour were used in testing. Driving postures were recorded using a sonic digitizer to measure the 3D locations of body landmarks. All test variables had significant independent effects on driving posture. Drivers were found to adapt to changes in the vehicle geometry primarily by changes in limb posture, whereas torso posture remained relatively constant. Stature accounts for most of the anthropometrically related variability in driving posture, and gender differences appear to be explained by body size variation. Large intersubject differences in torso posture, which are fairly stable across different seat and package conditions, are not closely related to standard anthropometric measures. The findings can be used to predict the effects of changes in vehicle and seat design on driving postures for populations with a wide range of anthropometric characteristics.

Time-to-Collision Judgments Under Realistic Driving Conditions

Objective: This study examined perceived time to collision (TTC) with automobile drivers under realistic approach, rear-end crash scenario conditions. Background: TTC refers to the time before impact if prevailing conditions continue. Method: In this test track study involving 51 drivers ranging from 20 to 70 years old, the drivers vision was occluded at either 3.6 or 5.6 s TTC during an in-lane approach to a lead vehicle. Drivers provided TTC estimates by pressing a button the instant they felt that they would have collided with the vehicle ahead. Results: Results indicated that TTC was consistently underestimated. The TTC ratio (perceived TTC/actual TTC) increased as driver speed decreased and as relative speed increased. These ratios were largely unaffected by age, gender, actual TTC, viewing time (1 s vs. continuous), and the presence of an eyes-forward, mental addition distraction task. Conclusion: Overall, these results suggest that under these low TTC conditions drivers estimate TTC in a relatively uniform fashion and that they are capable of providing this estimate based on a brief glimpse to the vehicle ahead. Application: These results are being used to develop an alert timing approach for a forward collision warning system intended to assist drivers in avoiding rear-end crashes with the vehicle ahead.

DEVELOPMENT OF AN IMPROVED DRIVER EYE POSITION MODEL

Society of Automotive Engineers (SAE) Recommended Practice J941 describes the eyellipse, a statistical representation of driver eye locations. Eye position data collected recently at University of Michigan Transportation Research Institute (UMTRI) suggest that the SAE J941 practice could be improved. SAE J941 currently uses the SgRP (vehicle seating reference point) location, seat track travel (L23), and design seatback angle (L40) as inputs to the eyellipse model. However, UMTRI data show that the characteristics of empirical eyellipses can be predicted more accurately using seat height, steering wheel position, and seat track rise. A series of UMTRI studies collected eye location data from groups of 50 to 120 drivers with statures spanning over 97 percent of the United States population. Data were collected in thirty-three vehicles. Significant and consistent differences were observed between eye position data collected before and after driving, indicating that actual driving is an important protocol feature for accurate measurement of driver eye position. In six vehicles, eyellipses obtained with two-way and six-way seat track travel were only slightly different. On average, drivers select seatback angles that are about 1.6 degrees more upright than design seatback angles. Stepwise regression techniques were used to identify the vehicle variables that have important effects on the distribution of driver eye locations. (A) For the covering abstract of the conference see IRRD 492369.

Anthropometric and Postural Variability: Limitations of the Boundary Manikin Approach

Human figure models are commonly used to facilitate ergonomic assessments of vehicle driver stations and other workplaces. One routine method of workstation assessment is to conduct a suite of ergonomic analyses using a family of boundary manikins, chosen to represent a range of anthropometric extremes on several dimensions. The suitability of the resulting analysis depends both on the methods by which the boundary manikins are selected and on the methods used to posture the manikins. The automobile driver station design problem is used to examine the relative importance of anthropometric and postural variability in ergonomic assessments. Postural variability is demonstrated to be nearly as important as anthropometric variability when the operator is allowed a substantial range of component adjustment. The consequences for boundary manikin procedures are discussed, as well as methods for conducting accurate and complete assessments using the available tools.

Comparing the effects of age, BMI and gender on severe injury (AIS 3+) in motor-vehicle crashes

BACKGROUND The effects of age, body mass index (BMI) and gender on motor vehicle crash (MVC) injuries are not well understood and current prevention efforts do not effectively address variability in occupant characteristics. OBJECTIVES (1) Characterize the effects of age, BMI and gender on serious-to-fatal MVC injury. (2) Identify the crash modes and body regions where the effects of occupant characteristics on the numbers of occupants with injury is largest, and thereby aid in prioritizing the need for human surrogates that represent different types of occupant characteristics and adaptive restraint systems that consider these characteristics. METHODS Multivariate logistic regression was used to model the effects of occupant characteristics (age, BMI, gender), vehicle and crash characteristics on serious-to-fatal injuries (AIS 3+) by body region and crash mode using the 2000-2010 National Automotive Sampling System (NASS-CDS) dataset. Logistic regression models were applied to weighted crash data to estimate the change in the number of annual injured occupants with AIS 3+ injury that would occur if occupant characteristics were limited to their 5th percentiles (age≤17 years old, BMI≤19kg/m(2)) or male gender. RESULTS Limiting age was associated with a decrease in the total number of occupants with head [8396, 95% CI 6871-9070] and thorax injuries [17,961, 95% CI 15,960-18,859] across all crash modes, decreased occupants with spine [3843, 95% CI 3065-4242] and upper extremity [3578, 95% CI 1402-4439] injuries in frontal and rollover crashes and decreased abdominal [1368, 95% CI 1062-1417] and lower extremity [4584, 95% CI 4012-4995] injuries in frontal impacts. The age effect was modulated by gender with older females more likely to have thorax and upper extremity injuries than older males. Limiting BMI was associated with 2069 [95% CI 1107-2775] fewer thorax injuries in nearside crashes, and 5304 [95% CI 4279-5688] fewer lower extremity injuries in frontal crashes. Setting gender to male resulted in fewer occupants with head injuries in farside crashes [1999, 95% CI 844-2685] and fewer thorax [5618, 95% CI 4212-6272], upper [3804, 95% CI 1781-4803] and lower extremity [2791, 95% CI 2216-3256] injuries in frontal crashes. Results indicate that age provides the greater relative contribution to injury when compared to gender and BMI, especially for thorax and head injuries. CONCLUSIONS Restraint systems that account for the differential injury risks associated with age, BMI and gender could have a meaningful effect on injury in motor-vehicle crashes. Computational models of humans that represent older, high BMI, and female occupants are needed for use in simulations of particular types of crashes to develop these restraint systems.

Automobile Occupant Posture Prediction for Use with Human Models

A new method of predicting automobile occupant posture is presented. The Cascade Prediction Model approach combines multiple independent predictions of key postural degrees of freedom with inverse kinematics guided by data-based heuristics. The new model, based on posture data collected in laboratory mockups and validated using data from actual vehicles, produces accurate posture predictions for a wide range of passenger car interior geometries. Inputs to the model include vehicle package dimensions, seat characteristics, and occupant anthropometry. The Cascade Prediction Model was developed to provide accurate posture prediction for use with any human CAD model, and is applicable to many vehicle design and safety assessment applications.

ATD POSITIONING BASED ON DRIVER POSTURE AND POSITION

Current automotive dynamic testing (ATD) positioning practices depend on seat track position, seat track travel range, and design seatback angle to determine appropriate occupant position and orientation for impact testing. In a series of studies conducted at the University of Michigans Transportation Research Institute, driver posture and position data were collected in 44 vehicles. Seat track reference points presently used to position ATDs were found to be poor predictors of the average seat positions selected by small female, midsize male, and large male drivers. Driver-selected seatback angle was not closely related to design seatback angle, the measure currently used to orient the ATD torso. A new ATD Positioning Model was developed that more accurately represents the seated posture and position of drivers who match the ATD statutes. Seat position is specified for each adult ATD size to match the mean predicted seat position of drivers matching the ATD reference stature. ATD torso orientation is set to the average driver torso orientation. The new positioning model places the ATDs in postures/positions that are more representative of drivers of similar size.