No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Needs of Autonomous Vehicles for Safe Operation on Existing Highways
Abstract
Autonomous vehicles (AVs) will gradually supersede human-driven vehicles (HDVs) in the future. Unlike human drivers who perceive their surroundings with their eyes, AV senses the ambient environments based on sensor fusion. More specifically, light detection and ranging (Lidar) sensors, video cameras, and radars are combined to help AV understand their surroundings. Current highway geometric design elements are mainly based on human driver-related parameters, such as perception and reaction time (PRT). However, AV will mix with HDV on existing highway infrastructures during the transition period. This study focuses on the sight distance aspect of highway geometric design. It is unsafe if an AV cannot effectively detect the objects within its required sight distance using the fused sensing system. Therefore, determining the needed sensor configurations (e.g. height, effective range, and field of view) is necessary for safe AV operation. This study used a simulation approach to determine the needed sensor configurations. First, the required stopping, decision, and passing sight distances for AV were determined considering that AV has a much shorter PRT than HDV. Second, the automated driving toolbox of MATLAB was used to construct different scenarios, each involving the road, obstacles, and actors. The road models were created following current design guides. A virtual AV equipped with a Lidar, cameras, radars, and an impeding agent served as the main actors on each road model. Third, many simulations were conducted with different sensor configurations to determine the Lidar configuration that achieves 100% detection for the safe operation of AV on existing highways.
ResearchGate has not been able to resolve any citations for this publication.
The three main elements of autonomous vehicles (AV) are visibility, orientation, and decision. This chapter presents an overview of the implementation of visibility-based technologies and methodologies. The chapter first presents two fundamental aspects that are necessary for understanding the main contents. The first aspect is highway geometric design as it relates to sight distance and highway alignment. The second aspect is mathematical basics, including coordinate transformation and visual space segmentation. Details on the Light Detection and Ranging (Lidar) system, which represents the ‘eye’ of the AV are presented. In particular, a new LiDAR 3D mapping system that can be operated on different platforms and modes for a new mapping scheme is described. The visibility methodologies include two types. Infrastructure visibility mainly addresses high-precision maps and sight obstacle detection. Traffic visibility (vehicles, pedestrians, and cyclists) addresses identification of critical positions and visibility estimation. Then, an overview of the decision element (path planning and intelligent car-following) for the movement of AV is presented. The chapter provides important information for researchers and therefore should help to advance road safety for autonomous vehicles.
The paper proposes a method supported by MATLAB for detection and measurement of missing point regions (MPR) which may cause severe road information loss in mobile laser scanning (MLS) point clouds. First, the scan-angle thresholds are used to segment the road area for MPR detection. Second, the segmented part is mapped onto a binary image with a pixel size of ε through rasterization. Then, MPR featuring connected 1-pixels are identified and measured via image processing techniques. Finally, the parameters regarding MPR in the image space are reparametrized in relation to the vehicle path recorded in MLS data for a better understanding of MPR properties on the geodetic plane. Tests on two MLS datasets show that the output of the proposed approach can effectively detect and assess MPR in the dataset. The ε parameter exerts a substantial influence on the performance of the method, and it is recommended that its value should be optimized for accurate MPR detections.
This paper investigates the potential changes in the geometric design elements in response to a fully autonomous vehicle fleet. When autonomous vehicles completely replace conventional vehicles, the human driver will no longer be a concern. Currently, and for safety reasons, the human driver plays an inherent role in designing highway elements, which depend on the driver’s perception-reaction time, driver’s eye height, and other driver related parameters. This study focuses on the geometric design elements that will directly be affected by the replacement of the human driver with fully autonomous vehicles. Stopping sight distance, decision sight distance, and length of sag and crest vertical curves are geometric design elements directly affected by the projected change. Revised values for these design elements are presented and their effects are quantified using a real-life scenario. An existing roadway designed using current AASHTO standards has been redesigned with the revised values. Compared with the existing design, the proposed design shows significant economic and environmental improvements, given the elimination of the human driver.
Autonomous vehicle technologies are receiving great attention with increasing demands for autonomy for both civilian and military purposes. In previous work (Mohamed et al., 2016), the recent developments in autonomous vehicles in the fields of advanced control, perception and motion planning techniques is surveyed. In this paper, the state of research w.r.t. autonomous vehicles from different perspectives will be described. The capability to integrate data and knowledge from different sensors are essential. In addition, advanced perception techniques and the capability to locate obstacles and targets are necessary to properly operate autonomous systems. Moreover, achieve reliable levels of performance by determining the faults and enabling the system to operate with these faults in mind. Fault tolerance is required to analysing the measured input/output signals of the system. This paper will briefly survey the recent developments in the field of autonomous vehicles from the perspectives of sensor fusion, computer vision, system identification and fault tolerance.
On sag vertical curves underpassing a structure, the fascia of the structure may cut the line of sight and limit the available sight distance. The current AASHTO model for symmetric sag curves with overpasses is intended for those that lie close to the point of vertical intersection on long sag curves. This paper presents an improved model that provides the exact minimum sight distance (Sm) for any conditions including non-centered overpasses and short sag curves, which are typical in urban areas. The model explicitly includes as variables the overpass location, the overpass width, and the configuration of the first and second grades of the sag vertical curve. The model is formulated by using mathematical optimization that involves an objective function and constraints. The objective function minimizes the available sight distance subject to constraints involving the geometry of the curve and the overpass. The model is spreadsheet based and is solved with powerful optimization software. Application of the model is illustrated with some examples and the results show that the AASHTO model may underestimate Sm by 10% to 30%. A closed-form model for Sm on single-arc asymmetric sag curves is also presented for Sm > L. The proposed model provides a friendly, flexible, and accurate tool for sight distance analysis and should be of interest to highway designers and practitioners.
Autonomous vehicle technologies are receiving more and more attention with increasing demands for autonomy and performance for both civilian and military purposes. In our previous paper Mohamed et al (2016), the recent developments in autonomous vehicles in the fields of advanced control, perception and motion planning techniques is surveyed. In this paper, the state of research with respect to autonomous vehicles from a number of different perspectives will be described. Among the many factors that affect the performance of these autonomous systems is the capability to integrate data and knowledge from different sensors, both of which are essential. In addition, advanced perception techniques, within the environment, and the capability to locate obstacles and targets are necessary in order to properly operate these systems. Moreover, autonomous vehicles can achieve reliable levels of performance by determining the faults and enabling the system to operate with these faults in mind. In order to achieve fault tolerance, often it is required to determine the mathematical model of the dynamic system by analysing the measured input and output signals of the system. This paper will briefly survey the recent developments in the field of autonomous vehicles from the perspectives of sensor fusion, computer vision, system identification and fault tolerance.
This paper provides an introduction to the history and concepts of connected and automated vehicle systems. Connected vehicle (CV) systems have been an important focal point for the intelligent transportation systems program for a while already, based on their ability to support a wide range of ITS applications and to knit vehicles and infrastructure elements into a well-integrated transportation system. Automated vehicle (AV) systems have had a longer and more turbulent history, with a recent upsurge of interest sparked by Google's initiative. The AV interests have had a strong element of technology push, but this paper shows how the AV systems can improve transportation system operations when they are combined with CV systems. The general-interest media and internet have tended to confound CV and AV systems with each other, but this paper disentangles them to explain their differences as well as their potential synergies.
A mathematical model is derived for describing the critical nature of the passing maneuver on two-lane highways. This model is based on the hypothesis that a critical position exists during the passing maneuver where the passing sight distance requirements to either complete or abort the pass are equal. At this point, the decision to complete the pass will provide the same head-on clearance to an opposing vehicle as will the decision to abort the pass. Current highway practice in both designing and marking for passing sight distance uses a model that assumes that once a driver starts a pass, he must continue until the pass is completed. In other words, the model assumes that the driver has no opportunity to abort the pass. Because this hypothesis is unrealistic, the model derived here is recommended for determining new passing sight distance requirements for both designing and marking passing zones. Suggested values are given for these requirements. A brief analysis is also presented of the sensitivity of passing sight distance requirements to vehicle length. This analysis shows that the effect of truck length is not as dramatic as previously reported in the literature.
Considerable research effort has been devoted within the past 15 years to automating the driving of highway vehicles in order to improve their safety and efficiency of operation and to help to reduce traffic congestion. Although the highway environment is in some ways more structured than other environments in which automated vehicles have been proposed to operate, the density and complexity of road traffic still make the sensing and control problems challenging. Because highway vehicles are not 'unmanned' but are expected to carry passengers and to coexist with other passenger-carrying vehicles, the reliability and safety considerations in the design of their control systems are much more important than they are for vehicles that are truly unmanned. This paper reviews the progress that has been made in recent research on highway vehicle automation and indicates the important research challenges that still need to be addressed before highway automation can become an everyday reality.
This paper discusses swerving as an alternative for stopping as a primary obstacle avoidance maneuver. By swerving instead of stopping, we show it is possible to achieve significantly higher-speeds given the same sensor capability. We have demonstrated navigation at speeds which rely on the principles described in this paper
Several models have been developed to determine the minimum passing sight distance required for safe and efficient operation on two-lane highways. The American Association of State Highway and Transportation Officials has developed a model assuming that once the driver begins a pass, he/she has no opportunity but to complete it. This assumption is believed to result in exaggerated passing sight distance requirements. Considerably shorter passing sight distance values are presented in the Manual of Uniform Traffic Control Devices and are used as the marking standards in Canada and the U.S.A. More appropriate models have been developed considering the driver's opportunity to abort the pass, and are based on a critical sight distance which produces the same factor of safety whether the pass is completed or aborted. However, these models need to be revised to determine the passing sight distance requirements more accurately and to closely match field observations. In this paper, a revised model for determining the minimum required passing sight distance was developed, based on the concept of critical sight distance and considering the kinematic interaction between the passing, passed, and opposing vehicles. The results of the revised model were compared with field data and showed that the revised model simulates the passing manoeuvre better than the currently-available models which are either too conservative or too liberal. The results showed that the passing sight distance requirements recommended in the Manual of Uniform Traffic Control Devices are sufficient at low design speeds (50-60 k.p.h.) and for manoeuvres involving passenger cars only. For higher design speeds, the Manual of Uniform Traffic Control Devices standards are less than the passing sight distance required for safe and comfortable passes. The deficiency was found to increase with the increase in design speed, and reaches about 36% at a 120-k.p.h. design speed. Based on these results, major revisions to the current Manual of Uniform Traffic Control Devices marking standards are recommended.
Driving scenario designer
- Matworks
User’s manual and programming guide: high definition LiDAR sensor. HDL-64E S3
- Velodyne