VTechWorks Repository :: Browsing by Author "Doerzaph, Zachary R."

Browsing by Author "Doerzaph, Zachary R."

Now showing 1 - 20 of 51

Application of Proximity Sensors to In-vehicle Data Acquisition Systems
Krothapalli, Ujwal; Stowe, Loren; Doerzaph, Zachary R.; Petersen, Andy (National Surface Transportation Safety Center for Excellence, 2018-05-02)
Naturalistic driving studies rely on human data reductionists to manually review and annotate driving behaviors. This work is time-consuming, and algorithms that could scan and categorize video data could make the data reduction process faster and more efficient. This report describes research to develop pose estimation methods that can be applied to drivers in naturalistic settings. Three methods were explored: (1) a depth-sensor-based pose estimation; (2) a deformable parts-based model; and (3) a tiny-image-based driver activity classifier. The tiny-image-based approach was chosen as the final solution and tested using the VTTIMLP01 dataset, a collection of about 80,000 images from 25 participants in naturalistic driving and simulated naturalistic driving conditions. The model was applied to approximately 50,000 images from the dataset covering seven activity classes: Eating/Drinking, Talking, Visor, Center Stack, Texting, One Hand on the Wheel, and Both Hands on the Wheel. The model, without any aspect ratio changes to the input image, was able to predict the activity classes with an overall 70% accuracy. To obtain better accuracies for individual activity classes, a separate model was built for each class, which resulted in a model with an overall accuracy of 74%. The Texting class had the poorest class accuracy (56%) due to the foreshortening effect on the limbs in the given camera angle. The One Hand on the Wheel class had the best class accuracy (96%).
Applications of Event Data Recorder Derived Crash Severity Metrics to Injury Prevention
Dean, Morgan Elizabeth (Virginia Tech, 2023-05-25)
Since 2015, there have been more than 35,000 fatalities annually due to crashes on United States roads [1], [2]. Typically, road departure crashes account for less than 10% of all annual crash occupants yet comprise nearly one third of all crash fatalities in the US [3]. In the year 2020, road departure crashes accounted for 50% of crash fatalities [2]. Road departure crashes are characterized by a vehicle leaving the intended lane of travel, departing the roadway, and striking a roadside object, such as a tree or pole, or roadside condition, such as a slope or body of water. One strategy currently implemented to mitigate these types of crashes is the use of roadside barriers. Roadside barriers, such as metal guardrails, concrete barriers, and cable barriers, are designed to reduce the severity of road departure crashes by acting as a shield between the departed vehicle and more hazardous roadside obstacles. Much like new vehicles undergo regulatory crash tests, barriers must adhere to a set of crash test procedures to ensure the barriers perform as intended. Currently, the procedures for full-scale roadside barrier crash tests used to evaluate the crash performance of roadside safety hardware are outlined in The Manual for Assessing Safety Hardware (MASH) [4]. During roadside barrier tests, the assessment of occupant injury risk is crucial, as the purpose of the hardware is to prevent the vehicle from colliding with a more detrimental roadside object, all the while minimizing, and not posing additional, risk to the occupants. Unlike the new vehicle regulatory crash tests conducted by the National Highway Traffic Safety Administration (NHTSA), MASH does not require the use of instrumented anthropomorphic test devices (ATD). Instead, one of the prescribed occupant risk assessment methods in MASH is the flail space model (FSM), which was introduced in 1981 and models an occupant as an unrestrained point mass. The FSM is comprised of two crash severity metrics that can be calculated using acceleration data from the test vehicle. Each metric is prescribed a maximum threshold in MASH and if either threshold is exceeded during a crash test the test fails due to high occupant injury risk. Since the inception of the FSM metrics and their thresholds, the injury prediction capabilities of these metrics have only been re-investigated in the frontal crash mode, despite MASH prescribing an oblique 25-degree impact angle for passenger vehicle barrier tests. The focus of this dissertation was to use EDR data from real-world crashes to assess the current relevance of roadside barrier crash test occupant risk assessment methods to the modern vehicle fleet and occupant population. Injury risk prediction models were constructed for the two FSM-based metrics and five additional crash severity metrics for three crash modes: frontal, side, and oblique. For each crash mode and metric combination, four injury prediction models were constructed: one to predict probability of injury to any region of the body and three to predict probability of injury to the head/face, neck, and thorax regions. While the direct application of these models is to inform future revisions of MASH crash test procedures, the developed models have valuable applications for other areas of transportation safety besides just roadside safety. The final two chapters of this dissertation explore these additional applications: 1) assessing the injury mitigation effectiveness of an advanced automatic emergency braking system, and 2) informing speed limit selection that supports the safe system approach. The findings in this dissertation indicate that both the FSM and additional crash severity metrics do a reasonable job predicting occupant injury risk in oblique crashes. One of the additional metrics performs better than the two FSM metrics. Additionally, several occupant factors, such as belt status and age, play significant roles in occupant risk prediction. These findings have important implications for future revisions of MASH, which could benefit from considering additional metrics and occupant factors in the occupant risk assessment procedures.
Assessing Alternate Approaches for Conveying Automated Vehicle Intentions
Basantis, Alexis Rae (Virginia Tech, 2019-10-30)
Objectives: Research suggests the general public has a lack of faith in highly automated vehicles (HAV) stems from a lack of system transparency while in motion (e.g., the user not being informed on roadway perception or anticipated responses of the car in certain situations). This problem is particularly prevalent in public transit or ridesharing applications, where HAVs are expected to debut, and when the user has minimal training on, and control over, the vehicle. To improve user trust and their perception of comfort and safety, this study aimed to develop more detailed and tailored human-machine interfaces (HMI) aimed at relying automated vehicle intended actions (i.e., "intentions") and perceptions of the driving environment to the user. Methods: This project developed HMI systems, with a focus on visual and auditory displays, and implemented them into a HAV developed at the Virginia Tech Transportation Institute (VTTI). Volunteer participants were invited to the Smart Roads at VTTI to experience these systems in real-world driving scenarios, especially ones typically found in rideshare or public transit operations. Participant responses and opinions about the HMIs and their perceived levels of comfort, safety, trust, and situational awareness were captured via paper-based surveys administered during experimentation. Results: There was a considerable link found between HMI modality and users' reported levels of comfort, safety, trust, and situational awareness during experimentation. In addition, there were several key behavioral factors that made users more or less likely to feel comfortable in the HAV. Conclusions: Moving forward, it will be necessary for HAVs to provide ample feedback to users in an effort to increase system transparency and understanding. Feedback should consistently and accurately represent the driving landscape and clearly communicate vehicle states to users.
Assessing Alternate Approaches for Conveying Automated Vehicle ‘Intentions’
Basantis, Alexis; Miller, Marty; Doerzaph, Zachary R.; Neurauter, Luke (SAFE-D: Safety Through Disruption National University Transportation Center, 2020-05)
One of the biggest highly automated vehicle (HAV) market barriers may be a lack of user trust in the automated driving system itself. Research has shown that this lack of faith in the system primarily stems from a lack of system transparency while the vehicle is in motion—users are not informed how the car will react in an upcoming scenario—and not having an effective way to control the vehicle in the event of a system failure. This problem is particularly prevalent in public transit or ridesharing applications, where HAVs are expected to first appear and where the user has less training on and control over the vehicle. To improve user trust and perceptions of comfort and safety, this study evaluated human-machine interface (HMI) systems, focused on visual and auditory displays, to better relay the perceived driving environment and the automated vehicle “intentions” to the user. These HMI systems were then implemented into a HAV developed at the Virginia Tech Transportation Institute and tested with volunteer participants on the Smart Roads.
Assessing Effects of Object Detection Performance on Simulated Crash Outcomes for an Automated Driving System
Galloway, Andrew Joseph (Virginia Tech, 2023-07-11)
Highly Automated Vehicles (AVs) have the capability to revolutionize the transportation system. These systems have the possibility to make roads safer as AVs do not have limitations that human drivers do, many of which are common causes of vehicle crashes (e.g., distraction or fatigue) often defined generically as human error. The deployment of AVs is likely to be very gradual however, and there will exist situations in which the AV will be driving in close proximity with human drivers across the foreseeable future. Given the persistent crash problem in which the makority of crashes are attributed to driver error, humans will continue to create potential collision scenarios that an AV will be expected to try and avoid or mitigate if developed appropriately. The absence of unreasonable risk in an AVs ability to comprehend and react in these situations is referred to as operational safety. Unlike advanced driver assistance systems (ADAS), highly automated vehicles are required to perform the entirety of the dynamic driving task (DDT) and have a greater responsibility to achieve a high level of operational safety. To address this concern, scenario-based testing has increasingly become a popular option for evaluating AV performance. On a functional level, an AV typically consists of three basic systems: the perception system, the decision and path planning system, and vehicle motion control system. A minimum level of performance is needed in each of these functional blocks to achieve an adequate level of operational safety. The goal of this study was to investigate the effects that perception system performance (i.e., target object state errors) has on vehicle operational safety in collision scenarios similar to that created by human drivers. In the first part of this study, recent annual crash data was used to define a relevant crash population of possible scenarios involving intersections that an AV operating as an urban taxi may encounter. Common crash maneuvers and characteristics were combined to create a set of testing scenarios that represent a high iii percentage of the overall crash population. In the second part of this study, each test scenario was executed using an AV test platform during closed road testing to determine possible real-world perception system performance. This provided a measure of the error in object detection measurements compared to the ideal (i.e., where a vehicle was detected to be compared to where it actually was). In the third part of this study, a set of vehicle simulations were performed to assess the effect of perception system performance on crash outcomes. This analysis simulated hypothetical crashes between an AV and one other collision partner. First an initial worst-case impact configuration was defined and was based on injury outcomes seen in crash data. The AV was then simulated to perform a variety of evasive maneuvers based on an adaptation of a non-impaired driver model. The impact location and orientation of the collision partner was simulated as two states: one based on the object detection of an ideal perception system and the other based on the object detection of the perception system from the AV platform used during the road testing. For simulations in which the two vehicles contacted each other, a planar momentum-impulse model was used for impact modeling and injury outcomes were predicted using an omni-directional injury model taken from recent literature. Results from this study indicate that errors in perception system measurements can change the perceived occupant injury risk within a crash. Sensitivity was found to be dependent on the specific crash type as well as what evasive maneuver is taken. Sensitivities occurred mainly due to changes in the principal direction of force for the crash and the interaction within the injury risk prediction curves. In order to achieve full operational safety, it will likely be important to understand the influence that each functional system (perception, decision, and control) may have on AV performance in these crash scenarios.
Assessment of Psychophysiological Characteristics of Drivers Using Heart Rate from SHRP2 Face Video Data
Sarkar, Abhijit; Doerzaph, Zachary R.; Abbott, A. Lynn (2014-08-25)
The goal is to

Extract heart rate from face video

Understand the behavior of driver, e.g. cognitive load, panic attack, drowsiness, DUI

Develop automatic video reduction technique

Devise a tool for future
Cardiac Signals: Remote Measurement and Applications
Sarkar, Abhijit (Virginia Tech, 2017-08-25)
The dissertation investigates the promises and challenges for application of cardiac signals in biometrics and affective computing, and noninvasive measurement of cardiac signals. We have mainly discussed two major cardiac signals: electrocardiogram (ECG), and photoplethysmogram (PPG). ECG and PPG signals hold strong potential for biometric authentications and identifications. We have shown that by mapping each cardiac beat from time domain to an angular domain using a limit cycle, intra-class variability can be significantly minimized. This is in contrary to conventional time domain analysis. Our experiments with both ECG and PPG signal shows that the proposed method eliminates the effect of instantaneous heart rate on the shape morphology and improves authentication accuracy. For noninvasive measurement of PPG beats, we have developed a systematic algorithm to extract pulse rate from face video in diverse situations using video magnification. We have extracted signals from skin patches and then used frequency domain correlation to filter out non-cardiac signals. We have developed a novel entropy based method to automatically select skin patches from face. We report beat-to-beat accuracy of remote PPG (rPPG) in comparison to conventional average heart rate. The beat-to-beat accuracy is required for applications related to heart rate variability (HRV) and affective computing. The algorithm has been tested on two datasets, one with static illumination condition and the other with unrestricted ambient illumination condition. Automatic skin detection is an intermediate step for rPPG. Existing methods always depend on color information to detect human skin. We have developed a novel standalone skin detection method to show that it is not necessary to have color cues for skin detection. We have used LBP lacunarity based micro-textures features and a region growing algorithm to find skin pixels in an image. Our experiment shows that the proposed method is applicable universally to any image including near infra-red images. This finding helps to extend the domain of many application including rPPG. To the best of our knowledge, this is first such method that is independent of color cues.
Characterizing and Comparing the ADS Maneuver Execution Subsystem Performance of Two Vehicles
Gopiao, Joseph Brandon Bueno (Virginia Tech, 2023-06-07)
Automated driving systems (ADS) are projected to bring a plethora of benefits to society, such as enhanced road safety and heightened quality of life. However, placing one's trust in the hands of an automated system is still a large concern to society. To facilitate the large-scale adoption of ADSs, they must be stringently tested and evaluated prior to their deployment on public roadways due to their direct impact on the safety of other motorists and vulnerable road users. Currently, no standardized method of quantifying ADS performance exists, so this research project contributes to the evaluation of ADSs by developing and demonstrating a test method that solely characterizes the motion control subsystem of an ADS. The developed test method involved generating representative driving scenarios that exercised both the longitudinal and lateral control elements of an ADS. This method was then demonstrated using two test vehicles with different control system architectures by (1) defining and injecting a ground truth trajectory into the ADS, (2) characterizing the motion control subsystem by quantifying its ability to follow the ground truth path under both nominal conditions and conditions where disturbances were introduced, and (3) analyzing the response of each vehicle to characterize their respective control systems as well as identify differences between the two control architectures. First, a set of representative driving scenarios was created to test the longitudinal and lateral control elements both in isolation and in tandem. Multiple unique design variations of each scenario were created by implementing various target speeds, accelerations, and turning radii that map to both standard and emergency maneuvers. The parameters were set to match naturalistic driving or regulatory requirements identified as part of a literature review. Next, a reference trajectory—the ground truth set of waypoints that define the position and speed of the ADS—was generated for each driving scenario. This reference trajectory was implemented using three methods: recording the waypoint trail of a human driver and creating a synthetic waypoint list mathematically or with CarMaker, a simulation platform for automobile testing (IPG Automotive 2021). Once this step was completed, the reference trajectory was inserted into the ADS to isolate the motion control system and facilitate a repeatable test input. When the test vehicle was under ADS control, the experimenter served as the designated fallback user so they could take control of the vehicle if necessary. Finally, a set of test metrics related to the operation of the ADS (lateral offset, heading error, speed error, longitudinal stop position error, and test completion percentage) were calculated using kinematic data to characterize each motion control system architecture. The analysis of the kinematic metrics for each test scenario demonstrated that the method could effectively evaluate the performance of ADS in various scenarios and highlight the strengths and weaknesses of each system. The control system of Vehicle A consistently lagged in throttle and brake actuation and rounded corners by turning early and with a larger cornering radius. This control system also could not exceed a lateral acceleration of 3.5 m/s2 when under ADS control and limited its yaw rate to keep the lateral acceleration below this level. Consequently, this limitation caused the vehicle to turn wide for radius and speed combinations with a lateral acceleration greater than 3.5 m/s2. On the other hand, the control system of Vehicle B consistently exhibited a small delay before turning and tended to overshoot lane changes at higher lateral accelerations. Regarding disturbances, only the road grade significantly affected the response of both vehicles.
Connected Motorcycle Crash Warning Interfaces
Song, Miao; McLaughlin, Shane B.; Doerzaph, Zachary R. (Connected Vehicle/Infrastructure University Transportation Center (CVI-UTC), 2016-01-15)
Crash warning systems have been deployed in the high-end vehicle market segment for some time and are trickling down to additional motor vehicle industry segments each year. The motorcycle segment, however, has no deployed crash warning system to date. With the active development of next generation crash warning systems based on connected vehicle technologies, this study explored possible interface designs for motorcycle crash warning systems and evaluated their rider acceptance and effectiveness in a connected vehicle context. Four prototype warning interface displays covering three warning mode alternatives (auditory, visual, and haptic) were designed and developed for motorcycles. They were tested on-road with three connected vehicle safety applications - intersection movement assist, forward collision warning, and lane departure warning - which were selected according to the most impactful crash types identified for motorcycles. It showed that a combination of warning modalities was preferred to a single display by 87.2% of participants and combined auditory and haptic displays showed considerable promise for implementation. Auditory display is easily implemented given the adoption rate of in-helmet auditory systems. Its weakness of presenting directional information in this study may be remedied by using simple speech or with the help of haptic design, which performed well at providing such information and was also found to be attractive to riders. The findings revealed both opportunities and challenges of visual displays for motorcycle crash warning systems. More importantly, differences among riders of three major motorcycle types (cruiser, sport, and touring) in terms of riders’ acceptance of a crash warning interface were revealed. Based on the results, recommendations were provided for an appropriate crash warning interface design for motorcycles and riders in a connected vehicle environment.
Connected Motorcycle System Performance
Viray, Reginald; Noble, Alexandria M.; Doerzaph, Zachary R.; McLaughlin, Shane B. (Connected Vehicle/Infrastructure University Transportation Center (CVI-UTC), 2016-01-15)
This project characterized the performance of Connected Vehicle Systems (CVS) on motorcycles based on two key components: global positioning and wireless communication systems. Considering that Global Positioning System (GPS) and 5.9 GHz Dedicated Short-Range Communications (DSRC) may be affected by motorcycle rider occlusion, antenna mounting configurations were investigated. In order to assess the performance of these systems, the Virginia Tech Transportation Institute’s (VTTI) Data Acquisition System (DAS) was utilized to record key GPS and DSRC variables from the vehicle’s CVS Vehicle Awareness Device (VAD). In this project, a total of four vehicles were used where one motorcycle had a forward mounted antenna, another motorcycle had a rear mounted antenna, and two automobiles had centermounted antennas. These instrumented vehicles were then subject to several static and dynamic test scenarios on closed test track and public roadways to characterize performance against each other. Further, these test scenarios took into account motorcycle rider occlusion, relative ranges, and diverse topographical roadway environments. From the results, both rider occlusion and approach ranges were shown to have an impact on communications performance. In situations where the antenna on the motorcycle had direct lineof-sight with another vehicle’s antenna, a noticeable increase in performance can be seen in comparison to situations where the line of sight is occluded. Further, the forward-mounted antenna configuration provided a wider span of communication ranges in open-sky. In comparison, the rear-mounted antenna configuration experienced a narrower communication range. In terms of position performance, environments where objects occluded the sky, such as deep urban and mountain regions, relatively degraded performance when compared to open sky environments were observed.
The Crash Injury Risk to Rear Seated Passenger Vehicle Occupants
Tatem, Whitney M. (Virginia Tech, 2020-01-22)
Historically, rear seat occupants have been at a lower risk of serious injury and fatality in motor vehicle crashes than their front seat counterparts. However, many passive safety advancements that have occurred over the past few decades such as advanced airbag and seatbelt technology primarily benefit occupants of the front seat. Indeed, safety for front seat occupants has improved drastically in the 21st century, but has it improved so much that the front seat is now safer than the rear? Today, rear-seated occupants account for 10% of all passenger vehicle fatalities. In this era focused on achieving zero traffic deaths, the safety of rear-seated occupants must be further addressed. This dissertation analyzed U.S. national crash data to quantify the risk of injury and fatality to rear-seated passenger vehicle occupants while accounting for the influence of associated crash, vehicle, and occupant characteristics such as crash severity, vehicle model year, and occupant age/sex. In rear impacts, the risk of moderate-to-fatal injury was greater for rear-seated occupants than their front-seated counterparts. In high-severity rear impact crashes, catastrophic occupant compartment collapse can occur and carries with it a great fatality risk. In frontal impacts, there is evidence that the rear versus front seat relative risk of fatality has been increasing in vehicle model years 2007 and newer. Rear-seated occupants often sustained serious thoracic, abdomen, and/or head injuries that are generally related to seatbelt use. Seatbelt pretensioners and load limiters – commonplace technology in the front seating positions – aim to mitigate these types of injuries but are rarely provided as standard safety equipment in the rear seats of vehicles today. Finally, in side impacts, injury and fatality risks to rear- and front-seated occupants are more similar than in the other crash modes studied, though disparities in protection remain, especially in near-side vehicle-to-vehicle crashes. Finally, this work projects great injury reduction benefits if a rear seat belt reminder system were to be widely implemented in the U.S. vehicle fleet. This dissertation presents a comprehensive investigation of the factors that contribute to rear-seated occupant injury and/or fatality through retrospective studies on rear, front, and side impacts. The overall goal of this dissertation is to better quantify the current risk of injury to rear-seated occupants under a variety of crash conditions, compare this to the current risk to front-seated occupants, and, when possible, identify how exactly injuries are occurring and ways in which they may be prevented in the future. The findings can benefit automakers who seek to improve the effectiveness of rear seat safety systems as well as regulatory agencies seeking to improve was vehicle tests targeting rear seat passenger vehicle safety.
Crowd-sourced Connected-vehicle Warning Algorithm using Naturalistic Driving Data
Noble, Alexandria M.; McLaughlin, Shane B.; Doerzaph, Zachary R.; Dingus, Thomas A. (2014-08-25)
A Dataset of Vehicle and Pedestrian Trajectories from Normal Driving and Crash Events in One Year of Virginia Traffic Camera Data
Bareiss, Max G. (Virginia Tech, 2023-06-07)
Traffic cameras are those cameras operated with the purpose of observing traffic, often streaming video in real-time to traffic management centers. These camera video streams allow transportation authorities to respond to traffic events and maintain situational awareness. However, traffic cameras also have the potential to directly capture crashes and conflicts, providing enough information to perform reconstruction and gain insights regarding causation and remediation. Beyond crash events, traffic camera video also offers an opportunity to study normal driving. Normal driver behavior is important for traffic planners, vehicle designers, and in the form of numerical driver models is vital information for the development of automated vehicles. Traffic cameras installed by state departments of transportation have already been placed in locations relevant to their interests. A wide range of driver behavior can be studied from these locations by observing vehicles at all times and under all weather conditions. Current systems to analyze traffic camera video focus on detecting when traffic events occur, with very little information about the specifics of those events. Prior studies into traffic event detection or reconstruction used 1-7 cameras placed by the researchers and collected dozens of hours of video. Crashes and other interesting events are rare and cannot be sufficiently characterized by camera installations of that size. The objective of this dissertation was to explore the utility of traffic camera data for transportation research by modeling and characterizing crash and non-crash behavior in pedestrians and drivers using a captured dataset of traffic camera video from the Commonwealth of Virginia, named the VT-CAST (Virginia Traffic Cameras for Advanced Safety Technologies) 2020 dataset. A total of 6,779,726 hours of traffic camera video was captured from live internet streams from December 17, 2019 at 4:00PM to 11:59PM on December 31, 2020. Video was analyzed by a custom R-CNN convolutional neural network keypoint detector to identify the locations of vehicles on the ground. The OpenPifPaf model was used to identify the locations of pedestrians on the ground. The location, pan, tilt, zoom, and altitude of each traffic camera was reconstructed to develop a mapping between the locations of vehicles and pedestrians on-screen and their physical location on the surface of the Earth. These physical detections were tracked across time to determine the trajectories on the surface of the Earth for each visible vehicle and pedestrian in a random sample of the captured video. Traffic camera video offers a unique opportunity to study crashes in-depth which are not police reported. Crashes in the traffic camera video were identified, analyzed, and compared to nationally representative datasets. Potential crashes were identified during the study interval by inspecting Virginia 511 traffic alerts for events which occurred near traffic cameras and impacted the flow of traffic. The video from these cameras was manually reviewed to determine whether a crash was visible. Pedestrian crashes, which did not significantly impact traffic, were identified from police accident reports (PARs) as a separate analysis. A total of 292 crashes were identified from traffic alerts, and six pedestrian crashes were identified from PARs. Road departure and rear-end crashes occurred in similar proportions to national databases, but intersection crashes were underrepresented and severe and rollover cases were overrepresented. Among these crashes, 32% of single-vehicle crashes and 50% of multi-vehicle crashes did not appear in the Virginia crash database. This finding shows promise for traffic cameras as a future data source for crash reconstruction, indicating traffic cameras are a capable tool to study unreported crashes. The safe operation of autonomous vehicles requires perception systems which make accurate short-term predictions of driver and pedestrian behavior. While road user behavior can be observed by the autonomous vehicles themselves, traffic camera video offers another potential information source for algorithm development. As a fixed roadside data source, these cameras capture a very large number of traffic interactions at a single location. This allows for detailed analyses of important roadway configurations across a wide range of drivers. To evaluate the efficacy of this approach, a total of 58 intersections in the VT-CAST 2020 dataset were sampled for driver trajectories at intersection entry, yielding 58,180 intersection entry trajectories. K-means clustering was used to group these trajectories into a family of 45 trajectory clusters. Likely as a function of signal phase, distinct groups of accelerating, constant speed, and decelerating trajectories were present. Accelerating and decelerating trajectories each occurred more frequently than constant speed trajectories. The results indicate that roadside data may be useful for understanding broad trends in typical intersection approaches for application to automated vehicle systems or other investigations; however, data utility would be enhanced with detailed signal phase information. A similar analysis was conducted of the interactions between drivers and pedestrians. A total of 35 crosswalks were identified in the VT-CAST 2020 dataset with sufficient trajectory information, yielding 1,488 trajectories of drivers interacting with pedestrians. K-means clustering was used to group these trajectories into a family of 16 trajectory clusters. Distinct groups of accelerating, constant speed, and decelerating trajectories were present, including trajectory clusters which described vehicles slowing down around pedestrians. Constant speed trajectories occurred the most often, followed by accelerating trajectories and decelerating trajectories. As with the prior investigation, this finding suggests that roadside data may be used in the development of driver-pedestrian interaction models for automated vehicles and other use cases involving a combination of pedestrians and vehicles. Overall, this dissertation demonstrates the utility of standard traffic camera data for use in traffic safety research. As evidence, there are already three current studies (beyond this dissertation) using the video data and trajectories from the VT-CAST 2020 dataset. Potential future studies include analyzing the mobile phone use of pedestrians, analyzing mid-block pedestrian crossings, automatically performing roadway safety assessments, considering the behavior of drivers following congested driving, evaluating the effectiveness of work zone hazard countermeasures, and understanding roadway encroachments.
Decision-adjusted driver risk predictive models using kinematics information
Mao, Huiying; Guo, Feng; Deng, Xinwei; Doerzaph, Zachary R. (Elsevier, 2021-06)
Accurate prediction of driving risk is challenging due to the rarity of crashes and individual driver heterogeneity. One promising direction of tackling this challenge is to take advantage of telematics data, increasingly available from connected vehicle technology, to obtain dense risk predictors. In this work, we propose a decision-adjusted framework to develop optimal driver risk prediction models using telematics-based driving behavior information. We apply the proposed framework to identify the optimal threshold values for elevated longitudinal acceleration (ACC), deceleration (DEC), lateral acceleration (LAT), and other model parameters for predicting driver risk. The Second Strategic Highway Research Program (SHRP 2) naturalistic driving data were used with the decision rule of identifying the top 1% to 20% of the riskiest drivers. The results show that the decision-adjusted model improves prediction precision by 6.3% to 26.1% compared to a baseline model using non-telematics predictors. The proposed model is superior to models based on a receiver operating characteristic curve criterion, with 5.3% and 31.8% improvement in prediction precision. The results confirm that the optimal thresholds for ACC, DEC and LAT are sensitive to the decision rules, especially when predicting a small percentage of high-risk drivers. This study demonstrates the value of kinematic driving behavior in crash risk prediction and the necessity for a systematic approach for extracting prediction features. The proposed method can benefit broad applications, including fleet safety management, use-based insurance, driver behavior intervention, as well as connected-vehicle safety technology development.
Development and Testing Of The iCACC Intersection Controller For Automated Vehicles
Zohdy, Ismail Hisham (Virginia Tech, 2013-10-28)
Assuming that vehicle connectivity technology matures and connected vehicles hit the market, many of the running vehicles will be equipped with highly sophisticated sensors and communication hardware. Along with the goal of eliminating human distracted driving and increasing vehicle automation, it is necessary to develop novel intersection control strategies. Accordingly, the research presented in this dissertation develops an innovative system that controls the movement of vehicles using cooperative cruise control system (CACC) capabilities entitled: iCACC (intersection management using CACC). In the iCACC system, the main assumption is that the intersection controller receives vehicle requests from vehicles and advises each vehicle on the optimum course of action by ensuring no crashes occur while at the same time minimizing the intersection delay. In addition, an innovative framework has been developed (APP framework) using the iCACC platform to prioritize the movements of vehicles based on the number of passengers in the vehicle. Using CACC and vehicle-to-infrastructure connectivity, the system was also applied to a single-lane roundabout. In general terms, this application is considered quite similar to the concept of metering single-lane entrance ramps. The proposed iCACC system was tested and compared to three other intersection control strategies, namely: traffic signal control, an all-way stop control (AWSC), and a roundabout, considering different traffic demand levels ranging from low to high levels of congestion (volume-to-capacity ration from 0.2 to 0.9). The simulated results showed savings in delay and fuel consumption in the order of 90 to 45 %, respectively compared to AWSC and traffic signal control. Delays for the roundabout and the iCACC controller were comparable. The simulation results showed that fuel consumption for the iCACC controller was, on average, 33%, 45% and 11% lower than the fuel consumption for the traffic signal, AWSC and roundabout control strategies, respectively. In summary, the developed iCACC system is an innovative system because of its ability to optimize/model different levels of vehicle automation market penetrations, weather conditions, vehicle classes/models, shared movements, roundabouts, and passenger priority. In addition, the iCACC is capable of capturing the heterogeneity of roadway users (cyclists, pedestrians, etc.) using a video detection technique developed in this dissertation effort. It is anticipated that the research findings will contribute to the application of automated systems, connected vehicle technology, and the future of driverless vehicle management. Finally, the public acceptability of the new advanced in-vehicle technologies is a challenging task and this research will provide valuable feedback for researchers, automobile manufacturers, and decision makers in making the case to introduce such systems.
Development of a Threat Assessment Algorithm for Intersection Collision Avoidance Systems
Doerzaph, Zachary R. (Virginia Tech, 2007-11-27)
Relative to other roadway segments, intersections occupy a small portion of the overall infrastructure; however, they represent the location for nearly 41 % of the annual automotive crashes in the United States. Thus, intersections are an inherently dangerous roadway element and a prime location for vehicle conflicts. Traditional safety treatments are effective at addressing certain types of intersection safety deficiencies; however, cumulative traffic data suggests these treatments do not address a large portion of the crashes that occur each year. Intersection Collision Avoidance Systems (ICAS) represent a new breed of countermeasures that focus on the types of crashes that have not been reduced with the application of traditional methods. Incursion systems, a subset of ICAS, are designed to specifically undertake crashes that are a result of the violation of a traffic control device. Intersection Collision Avoidance Systems to address Violations (ICAS-V) monitor traffic as it approaches the intersection through a network of in-vehicle sensors, infrastructure- mounted sensors, and communication equipment. A threat-assessment algorithm performs computations to predict the driver's intended intersection maneuver, based on these sensor inputs. If the system predicts a violation, it delivers a timely warning to the driver with the aim of compelling the driver to stop. This warning helps the driver to avoid a potential crash with adjacent traffic. The following dissertation describes an investigation of intersection approach behavior aimed at developing a threat assessment algorithm for stop-sign intersections. Data were collected at live intersections to gather infrastructure-based naturalistic vehicle approach trajectories. This data were compiled and analyzed with the goal of understanding how drivers approach intersections under various speeds and environmental conditions. Six stop-controlled intersection approaches across five intersections in the New River Valley, Virginia area were selected as the test sites. Data were collected from each site for at least two months, resulting in over sixteen total months of data. A series of statistical analysis techniques were applied to construct a set of threat assessment algorithms for stop-controlled intersections. These analyses identified characteristics of intersection approaches that suggested driver intent at the stop sign. Models were constructed to predict driver stopping intent based on measured vehicle kinematics. These models were thoroughly tested using simulation and evaluated with signal detection theory. The overall output of this work is a set of algorithms that may be integrated into an ICAS-V for on-road testing.
Effectiveness of Automatic Emergency Braking for Protection of Pedestrians and Bicyclists in the U.S.
Haus, Samantha Helen (Virginia Tech, 2021-11-16)
In the United States, there were 36,560 traffic-related fatalities in 2018, of which 20% were pedestrians, bicyclists, and other vulnerable road users (VRUs) [1]. Vulnerable road users are non-vehicle occupants who, because they are not enclosed in a vehicle, are at higher risk of injury in traffic crashes. While overall traffic fatalities in the US have been decreasing, pedestrian and bicyclist fatalities have been trending upward. Vehicle-based active safety features could avoid or mitigate crashes with VRUs, but are highly dependent on the ability to detect a VRU with enough time or distance. This work presents methods to examine the characteristics of vehicle-pedestrian and vehicle-bicycle crashes and near-crashes using a variety of data sources, assess the potential effectiveness of Automatic Emergency Braking (AEB) in avoiding and mitigating VRU crashes through modeling and simulation, and estimate the future benefits of AEB for VRU safety in the United States. Additionally, active safety features are most effective when behavior of VRUs can be anticipated, however, the behavior of pedestrians and bicyclists is notoriously unpredictable. Therefore, an approach to examine and categorize pedestrian behavior in response to near-crashes and crashes events is presented. Overall, findings suggest that AEB has great potential to avoid and mitigate collisions with pedestrians and bicyclists, but it cannot avoid all crashes even when an idealized AEB system is assumed. Most pedestrians and bicyclists were found to be visible for at least one second prior to the crash, but obstructions, the unpredictability of VRUs, and adverse weather/lighting conditions still pose challenges in avoiding and mitigating crashes with VRUs.
Effectiveness of Compensatory Vehicle Control Techniques Exhibited by Drivers after Arthroscopic Rotator Cuff Surgery
Metrey, Mariette Brink (Virginia Tech, 2023-07-10)
Current return-to-drive recommendations for patients following rotator cuff repair (RCR) surgery are not uniform due to a lack of empirical evidence relating driving safety and time-after-surgery. To address the limitations of previous work, Badger et al. (2022) evaluated, on public roads, the driving fitness of patients prior to RCR and at multiple post-operative timepoints. The goal of the Badger, et al. study was to make evidence-based return-to-drive recommendations in an environment with higher fidelity than that of a simulator and not subject to biases inherent to surveys. Badger et al., however, do not fully investigate the driving practices exhibited by subjects, overlooking the potential presence of compensatory driver behaviors. Further investigation of these behaviors through observation of direct driving techniques and practices over time can specifically answer how drivers may modify their behaviors to address a perceived state of impairment. Additionally, the degree of success in vehicle operation by comparing an ideal turn to the path taken by the driver allows for a level of quantification of the effectiveness of the compensatory techniques. Moreover, driver trajectories inferred from the vehicle Controller Area Network (CAN) metrics and from global positioning system (GPS) coordinates are contrasted with the ideal turn to assess minimum requirements for future sensors that are used to make these trajectory comparisons. This investigation leverages pre-existing data collected by the Virginia Tech Transportation Institute (VTTI) and Carilion Clinic as used in the analysis performed by Badger et al. (2022). RCR patients (n=27) executed the same prescribed driving maneuvers and drove the same route in a preoperative state and at 2-, 4-, 6-, and 12-weeks post operation. Behavioral data were annotated to extract key characteristics of interest and related them to relevant vehicle sensor readings. To construct vehicle paths, data was obtained from the on-board data acquisition system (DAS). Behavioral metrics considered the use of ipsilateral vehicle controls, performance of non-primary vehicle tasks, and steering techniques utilized to assess the impact of mobility restrictions due to sling use. Sling use was found to be a significant factor in use of the non-ipsilateral hand associated with the operative extremity (i.e., operative hand) on vehicle functions and, in particular, difficulty with the gear shifting control. Additionally, when considering the performance of non-primary vehicle tasks as assessed through a prescribed visor manipulation, sling use was not a significant factor for the task duration or completion of the task in a fluid motion. Sling use was, however, significant with respect to operative hand position prior to the completion of the visor manipulation: the operative hand was often not on the steering wheel prior to the visor maneuver. In addition, the operative hand was never used to manipulate the visor when the sling was worn. One-handed steering was also more frequent with the presence of the sling. Further behavioral analysis assessed the presence of compensatory behavior exhibited by subjects during periods in which impairment was perceived. Perceived impairment was observed as a function of the different experimental timepoints. Findings indicated a significant decrease in the lateral vehicle jerk during post-operative weeks 6 and 12. Significant differences, however, were not observed in body position alteration to avoid contact with the interior vehicle cabin, in over-the-shoulder checks, and in forward leans during yield and merge maneuvers. Regarding trajectory analysis, sling use did not produce a significant difference in the error metrics between the actual and ideal paths. In completion of turning maneuvers, however, operative extremity was significant for left turns, with greater error against the ideal path observed from those in the left operative cohort compared to those in the right operative cohort. For the right turn, however, operative extremity was not found to be a significant factor. In addition, the GPS data accuracy proved insufficient to support comparison against the ideal path. Overall, findings from this study provide metrics beyond those used in Badger, et al. that can be used in answering when it is safe for rotator cuff repair patients to return-to-drive. With the limited differences observed as a function of study timepoint and sling use, it is recommended that patients are able to safely return-to-drive at two weeks post-operation. If anything, results suggest that overcompensation, as inferred from observation of safer driving behaviors than normal, is present during some experimental timepoints, particularly post-operative week 2.
Effectiveness of Vehicle External Communication Toward Improving Vulnerable Road User Safe Behaviors: Considerations for Legacy Vehicles to Automated Vehicles of the Future
Rossi-Alvarez, Alexandria Ida (Virginia Tech, 2023-01-25)
Automated vehicles (AVs) will be integrated into our society at some point in the future, but when is still up for debate. An extensive amount of research is being completed to understand the communication methods between AVs and other road users sharing the environment to prepare for this future. Currently, researchers are working to understand how different forms of external communication on the AVs will impact vulnerable road user (VRU) interaction. However, within the last 10 years, VRU casualty rates have continued to rise for all classifications of VRUs. Unfortunately, there is no suggestion that pedestrian fatality rates will ever decrease without some intervention. This dissertation aims at understanding the impacts of eHMI across real-world, complex scenarios with AVs and how researchers can apply those future findings to improve VRUs' judgments to today. A series of studies evaluated the necessity and impact of eHMI on AV–VRU interaction, assessed how the visual components of eHMI influenced VRU crossing decisions, and how variations in a real-world environment (multiple vehicles and scenario complexity) impact crossing decision behavior. Two studies examined how eHMI will impact future interactions between AVs and VRUs. Specifically, to understand how to advance the design of these future devices to avoid unintended consequences that may result. Results from these studies found that the presence and condition of eHMI did not influence participants' willingness to cross. Participants primarily relied on the speed and distance of the vehicle to make their crossing decision. It was difficult for participants to focus on the eHMI when multiple vehicles competed for their attention. Participants typically prioritized their focus on the vehicle that was nearest and most detrimental to their crossing path. Additionally, the type of scenario caused participants to make more cautious crossing decisions. However, it did not influence their willingness to cross. The last study applied the learnings from the first two studies to a foundational perception study for current legacy vehicles. These results showed a significant increase in judgment accuracies with a display. Through analysis across overall conclusions from the 3 studies, five critical findings were identified when addressing eHMI and 3 design recommendations, which are discussed in the penultimate section of this work. The results of this dissertation indicate that eHMI improved VRUs' accuracy of perception of change in vehicle speed. eHMI did not significantly impact VRUs crossing decisions. However, the complexity of the traffic scenarios affected the level of caution participants exhibited in their crossing behavior.
Enhanced Camera
White, Elizabeth E.; Chilcott, Dan; Doerzaph, Zachary R. (National Surface Transportation Safety Center for Excellence, 2016-08-09)
This report describes testing equipment and procedures developed by the Center for Technology Development at the Virginia Tech Transportation Institute (VTTI) to evaluate critical attributes for cameras used in naturalistic driving research. A survey of VTTI researchers was conducted to determine the most important attributes and known issues from past naturalistic driving studies. The data collected were used to design and build a camera testing apparatus and a field of view (FOV) testing apparatus that allow the standardized testing of a number of important attributes, including image clarity, black and white versus color display, FOV, and the quality of the image under various lighting conditions. The camera testing apparatus also has the ability to detect infrared (IR) sources in cameras. In addition, this report includes a set of recommended methods for testing system resolution using the International Organization for Standardization (ISO) 12233 standard test target and image sharpness using free software or commercial software that would provide increased accuracy and decreased training times. Procedures are also described for testing environmental factors such as temperature range, temperature cycling, and immersion.

Browsing by Author "Doerzaph, Zachary R."

Results Per Page

Sort Options