Browse Topic: Driving automation

Items (302)
Find
Robust Behavioral Cloning for Autonomous Vehicles Using End-to-End Imitation Learning12-04-03-00238/19/2021
In this work, we present a lightweight pipeline for robust behavioral cloning of a human driver using end-to-end imitation learning. The proposed pipeline was employed to train and deploy three distinct driving behavior models onto a simulated vehicle. The training phase comprised of data collection, balancing, augmentation, preprocessing, and training a neural network, following which the trained model was deployed onto the ego vehicle to predict steering commands based on the feed from an onboard camera. A novel coupled control law was formulated to generate longitudinal control commands on the go based on the predicted steering angle and other parameters such as the actual speed of the ego vehicle and the prescribed constraints for speed and steering. We analyzed the computational efficiency of the pipeline and evaluated the robustness of the trained models through exhaustive experimentation during the deployment phase. We also compared our approach against state-of-the-art implementation in order to comment on its validity.
Samak, Tanmay Vilas, Samak, Chinmay Vilas, Kandhasamy, Sivanathan
Title updates jmsSAE-PP-004008/13/2021
Abstract
Jackson, Alyssa
Automated Driving System Data LoggerJ3197_202107 (Current)7/23/2021
This SAE Recommended Practice provides common data output formats and definitions for a variety of data elements that may be useful for analyzing the performance of automated driving system (ADS) during an event that meets the trigger threshold criteria specified in this document. The document is intended to govern data element definitions, to provide a minimum data element set, and to specify a common ADS data logger record format as applicable for motor vehicle applications. Automated driving systems (ADSs) perform the complete dynamic driving task (DDT) while engaged. In the absence of a human “driver,” the ADS itself could be the only witness of a collision event. As such, a definition of the ADS data recording is necessary in order to standardize information available to the accident reconstructionist. For this purpose, the data elements defined herein supplement the SAE J1698-1 defined EDR in order to facilitate the determination of the background and events leading up to a collision in an ADS-operated vehicle. The data elements defined in this document are unique to Level 3, 4, or 5 ADS features, as defined by SAE J3016, and provide additional background of the events leading up to a crash or crash-like event. The data from sensors such as camera(s), LiDAR(s) etc. will provide information in the absence of a human driver. The data included in the ADS data logger is expected to be used in conjunction with the SAE J1698 event data recorder (EDR) record and traditional accident reconstruction analysis. The EDR and ADS data logger will capture information leading up to the triggered event, at a minimum. There are no facts to support that recording data for greater than 5 seconds pre-event would change the outcome of any crash analysis. Thus, the recommended recording duration for a data logger is 5 seconds pre-event, same as an EDR. Due to the potential for sensor and/or communication failure during a crash event, the recommendation is that data should be collected post-crash for impact and rollover sensors for up to 250 ms. ADS technology is still being developed and is not yet commercially deployed. Therefore, this SAE Recommended Practice is intended as a guide toward standard practice and is subject to change to keep pace with experience and technical advances.
Event Data Recorder Committee
Taxonomy and Definitions for Terms Related to Cooperative Driving Automation for On-Road Motor VehiclesJ3216_202107 (Current)7/16/2021
This document describes machine-to-machine (M2M) communication to enable cooperation between two or more participating entities or communication devices possessed or controlled by those entities. The cooperation supports or enables performance of the dynamic driving task (DDT) for a subject vehicle with driving automation feature(s) engaged. Other participants may include other vehicles with driving automation feature(s) engaged, shared road users (e.g., drivers of manually operated vehicles or pedestrians or cyclists carrying personal devices), or road operators (e.g., those who maintain or operate traffic signals or workzones). Cooperative driving automation (CDA) aims to improve the safety and flow of traffic and/or facilitate road operations by supporting the movement of multiple vehicles in proximity to one another. This is accomplished, for example, by sharing information that can be used to influence (directly or indirectly) DDT performance by one or more nearby road users. Vehicles and infrastructure elements engaged in cooperative automation may share information, such as state (e.g., vehicle position, signal phase), intent (e.g., planned vehicle trajectory, signal timing), or seek agreement on a plan (e.g., coordinated merge). Cooperation among multiple participants and perspectives in traffic can improve safety, mobility, situational awareness, and operations. However, nothing in this document is intended to suggest that driving automation requires such cooperation in order to be performed safely. Cooperative strategies may be enabled by the sharing of information in a way that meets the needs of a given application. The needs may be expressed in terms of performance characteristics, such as latency, transmission mode (e.g., one-way, two-way), range, privacy and security, and information content and quality. There are several potential technologies for communicating information between the subject vehicle and other participants. This document focuses on application-oriented functionality and does not imply the need for or require any specific functionality associated with communications protocols or the open systems interconnection model layers in a protocol stack. This document addresses the operational and tactical timescales of dynamic driving on ADS-operated vehicles, and excludes strategic functions such as trip scheduling and selection of destinations and waypoints. This information report is intended to facilitate communication and awareness for the design and anticipated development and validation of cooperative driving automation.
Cooperative Driving Automation(CDA) Committee
AVSC Information Report for Adapting a Safety Management System (SMS) for Automated Driving System (ADS) SAE Level 4 and 5 Testing and EvaluationAVSC000072021077/16/2021
This AVSC Information Report for Adapting a Safety Management System (SMS) for Automated Driving System (ADS) SAE Level 4 and 5 Testing and Evaluation (AVSC000007202107) shares information on a Safety Management System (SMS) framework in the context of ADS testing and evaluation operations. This report was developed to provide ADS organizations information about the role of organizational safety, which may be considered in the ADS testing and evaluation process. The SMS framework represents a method used in non-automotive industries (e.g., aviation, rail, nuclear) with the goal of enhancing an organization’s operational safety performance. A Safety Management System (SMS) is one approach designed to support organizational safety in a systematic and integrated way. The SMS framework is intended to promote a positive safety culture, assess and manage safety risk, evaluate risk control effectiveness, and support organizational safety policies and objectives. This information report affords organizations another safety management approach to add to their ADS test operations.
Automated Vehicle Safety Consortium
Taxonomy and Definition of Safety Principles for Automated Driving System (ADS)J3206_202107 (Current)7/7/2021
This SAE Information Report classifies and defines a harmonized set of safety principles intended to be considered by ADS and ADS-equipped vehicle development stakeholders. The set of safety principles herein is based on the collection and analysis of existing information from multiple entities, reflecting the content and spirit of their efforts, including: SAE ITC AVSC Best Practices CAMP Automated Vehicle Research for Enhanced Safety - Final Report RAND Report - Measuring Automated Vehicle Safety: Forging a Framework U.S. DOT: Automated Driving Systems 2.0 - A Vision for Safety Safety First for Automated Driving (SaFAD) UNECE WP29 amendment proposal UNECE/TRANS/WP.29/GRVA/2019/13 On a Formal Model of Safe and Scalable Self-Driving Cars (Intel RSS model) SAE J3018 This SAE Information Report provides guidance for the consideration and application of the safety principles for the development and deployment of ADS and ADS-equipped vehicles. This SAE Information Report is not intended to encompass all aspects of system-level safety for an ADS-equipped vehicle, including communication with other traffic participants. Addressing all identified safety principles is intended to support, but not fully ensure, comprehensive system-level safety. As an SAE Information Report, this document is non-normative, imposes no requirements, and does not address: Requirements for methodology, metrics, and/or acceptance thresholds. Ethics-related safety principles, or any link between the safety principles defined in this document and ethical studies/frameworks. Conformance with safety principles for purposes of liability and/or fault assignment. As ADS technology and deployment are expanded in the future, this document may be reconsidered for future revision including normative requirements.
On-Road Automated Driving (ORAD) committee
Taxonomy and Definitions for Terms Related to Driving Automation Systems for On-Road Motor VehiclesJ3016_202104 (Current)4/30/2021
This document describes [motor] vehicle driving automation systems that perform part or all of the dynamic driving task (DDT) on a sustained basis. It provides a taxonomy with detailed definitions for six levels of driving automation, ranging from no driving automation (Level 0) to full driving automation (Level 5), in the context of [motor] vehicles (hereafter also referred to as “vehicle” or “vehicles”) and their operation on roadways: Level 0: No Driving Automation Level 1: Driver Assistance Level 2: Partial Driving Automation Level 3: Conditional Driving Automation Level 4: High Driving Automation Level 5: Full Driving Automation These level definitions, along with additional supporting terms and definitions provided herein, can be used to describe the full range of driving automation features equipped on [motor] vehicles in a functionally consistent and coherent manner. “On-road” refers to publicly accessible roadways (including parking areas and private campuses that permit public access) that collectively serve all road users, including cyclists, pedestrians, and users of vehicles with and without driving automation features. The levels apply to the driving automation feature(s) that are engaged in any given instance of on-road operation of an equipped vehicle. As such, although a given vehicle may be equipped with a driving automation system that is capable of delivering multiple driving automation features that perform at different levels, the level of driving automation exhibited in any given instance is determined by the feature(s) that are engaged. This document also refers to three primary actors in driving: the (human) user, the driving automation system, and other vehicle systems and components. These other vehicle systems and components (or the vehicle in general terms) do not include the driving automation system in this model, even though as a practical matter a driving automation system may actually share hardware and software components with other vehicle systems, such as a processing module(s) or operating code. The levels of driving automation are defined by reference to the specific role played by each of the three primary actors in performance of the DDT and/or DDT fallback. “Role” in this context refers to the expected role of a given primary actor, based on the design of the driving automation system in question and not necessarily to the actual performance of a given primary actor. For example, a driver who fails to monitor the roadway during engagement of a Level 1 adaptive cruise control (ACC) system still has the role of driver, even while s/he is neglecting it. Active safety systems, such as electronic stability control (ESC) and automatic emergency braking (AEB), and certain types of driver assistance systems, such as lane keeping assistance (LKA), are excluded from the scope of this driving automation taxonomy because they do not perform part or all of the DDT on a sustained basis, but rather provide momentary intervention during potentially hazardous situations. Due to the momentary nature of the actions of active safety systems, their intervention does not change or eliminate the role of the driver in performing part or all of the DDT, and thus are not considered to be driving automation, even though they perform automated functions. In addition, systems that inform, alert, or warn the driver about hazards in the driving environment are also outside the scope of this driving automation taxonomy, as they neither automate part or all of the DDT, nor change the driver’s role in performance of the DDT (see 8.13). It should be noted, however, that crash avoidance features, including intervention-type active safety systems, may be included in vehicles equipped with driving automation systems at any level. For automated driving system (ADS) features (i.e., Levels 3 to 5) that perform the complete DDT, crash mitigation and avoidance capability is part of ADS functionality (see also 8.13).
On-Road Automated Driving (ORAD) committee
Development of Fault Detection and Emergency Control for Application to Autonomous Vehicle2021-01-00754/6/2021
This paper describes a failsafe system of automated driving vehicles. The failsafe system consists of the following two parts: sliding mode observer-based environment sensor, chassis sensor fault detection, and emergency deceleration control. Two sliding mode observers are designed to reconstruct the fault of acceleration and environment sensor(Lidar) in a longitudinal direction. In the environment sensor's fault detection part, the longitudinal vehicle model receives clearance and relative velocity values. Therefore, failure diagnosis is possible regardless of environmental sensors, such as radar, lidar, and camera. This paper's sensor data is the failure of Delphi's Electronically Scanning Radar (ESR) and Ibeo's LUX Lidar installed in an autonomous vehicle. The emergency deceleration control algorithm employs the sliding mode control with adaptive convergence time. In the event of a failure, it is significant to control the vehicle within a short period safely. The Adaptive convergence time concept proves a mathematical convergence of the vehicle control time after a failure occurs. As soon as the error occurred, the error was proved to always converge to zero within the final time. Thus, the proposed method introduces the concept of convergence time, and mathematically demonstrates that the state reached the reference target within the specified time when a failure occurred. In the emergency control part, two processing unit hardware structures are adopted to comply with SAE International standard J3016 and NHTSA autonomous vehicle safety report standards. The proposed fail-safe detection algorithm is evaluated through vehicle test data, and the fail-safe control algorithm evaluates through computer simulation and vehicle tests.
Jong Min, Lee, Oh, Kwang Seok, Song, Taejun
Correlation between Sensor Performance, Autonomy Performance and Fuel-Efficiency in Semi-Truck Platoons2021-01-00644/6/2021
Semi-trucks, specifically class-8 trucks, have recently become a platform of interest for autonomy systems. Platooning involves multiple trucks following each other in close proximity, with only the lead truck being manually driven and the rest being controlled autonomously. This approach to semi-truck autonomy is easily integrated on existing platforms, reduces delivery times, and reduces greenhouse gas emissions via fuel economy benefits. Level 1 SAE fuel studies were performed on class-8 trucks operating with the Auburn Cooperative Adaptive Cruise Control (CACC) system, and fuel savings up to 10-12% were seen. Enabling platooning autonomy required the use of radar, global positioning systems (GPS), and wireless vehicle-to-vehicle (V2V) communication. Poor measurements and state estimates can lead to incorrect or missing positioning data, which can lead to unnecessary dynamics and finally wasted fuel. This is especially an issue if deceleration is applied in response to a bad measurement. In this study, a faulty radar was shown to cause a greater than 5% increase in fuel consumption. The mechanism of this fuel consumption increase is investigated and applied to other types of sensor failures to indicate their potential effects on fuel economy. This analysis indicates that poor GPS signals over short time can be largely filtered out, with no real gain or loss of fuel economy. V2V communications were intentionally limited by causing interference, which resulted in dropped communication packets over a small physical area, but not an appreciable impact on fuel economy.
Adam, Cristian, Lakshmanan, Sridhar, Richardson, Paul, Stegner, Evan, Ward, Jacob, Hoffman, Mark, Bevly, David M.
Driving Automation System Test Scenario Development Process Creation and Software-in-the-Loop Implementation2021-01-00624/6/2021
Automated driving systems (ADS) are one of the key modern technologies that are changing the way we perceive mobility and transportation. In addition to providing significant access to mobility, they can also be useful in decreasing the number of road accidents. For these benefits to be realized, candidate ADS need to be proven as safe, robust, and reliable; both by design and in the performance of navigating their operational design domain (ODD). This paper proposes a multi-pronged approach to evaluate the safety performance of a hypothetical candidate system. Safety performance is assessed through using a set of test cases/scenarios that provide substantial coverage of those potentially encountered in an ODD. This systematic process is used to create a library of scenarios, specific to a defined domain. Beginning with a system-specific ODD definition, a set of core competencies are identified. These core competencies are then considered both in isolation and in conjunction with other potential confounding factors (e.g. other traffic or atmospheric conditions); with “edge cases” being represented as compounded or unique sets of confounding factors. Using this approach, a candidate scenario set is presented, along with a discussion of nuances and necessary considerations in scenario selection. These approaches are combined in a simulated environment to demonstrate their use. Finally, a strategy is proposed to automate the overall scenario testing process to make the execution less cumbersome. This process of test scenario creation strictly follows the ISO 26262 concept phase to verify the safety goals and functional safety requirements.
Patil, Mayur, Lybarger, Alexander, Midlam-Mohler, Shawn, Stoddart, Evan
Techno-Economic Analysis of Fixed-Route Autonomous and Electric Shuttles2021-01-00614/6/2021
This paper takes a realistic approach to develop a techno-economic analysis for fixed-route autonomous shuttles. To develop a model for analysis, the current state of technology was used to approximate three timelines for achieving SAE level 5 capabilities: progressive, realistic, and conservative. Within these timelines, there are four different increments for advancements in the technology laid out as follows: SAE Level 0 - human driver, SAE Level 4 - in-vehicle safety operator, SAE Level 4 - remote safety operator, and SAE Level 5 - no safety operator. These increments in the changes of the technology were chosen based on the trends in the industry. Various shuttle models were used based on different rider quantities and drive-train requirements (electric vs gas) in this analysis. This allows for further understanding of how these deployment plans will vary the cost for shuttles operating in high, mid, and low ridership demand environments. Additional drive-train comparison shows the savings based on the choice of electric vs ICE vehicles. Taking these parameters into consideration, simulations were run for the various vehicle models in the various ridership demand environments to produce the economic costs for each situation. It was found that in 15 years there is an economic savings of 72%, 68%, 43%, and 35% for small, medium, large, and extra-large shuttles, respectively if deployed with the conservative plan of becoming SAE Level 4 with an in-vehicle safety operator in 2 years, SAE Level 4 with a remote safety operator in 4 years, and SAE Level 5 in 8 years.
Goberville, Nick, Zoardar, Md Marsad, Rojas, Johan, Brown, Nicolas, Motallebiaraghi, Farhang, Navarro, Anthony, Asher, Zachary
No Cost Autonomous Vehicle Advancements in CARLA through ROS2021-01-01064/6/2021
Development of autonomous vehicle technology is expensive and perhaps more complicated than initially thought, as evidenced by the recent rollback of anticipated delivery dates from companies such as Tesla, Waymo, GM, and more. One of the most effective techniques to reduce research and development costs and speed up implementation is rigorous analysis through simulation. In this paper, we present multiple autonomous vehicle perception and control strategies that are rigorously investigated in the user friendly, free, and open-source simulation environment, CARLA. Overall, we successfully formulated potential solutions to the autonomous navigation problem and assessed their advantages and disadvantages in simulation at no cost. First, a lane finding method utilizing polynomial fitting and machine learning is proposed. Then, the waypoint navigation strategy is described, along with route planning. Object detection is then implemented using pre-trained convolutional neural networks. A classic PID control strategy and the Stanley Method were investigated for lateral and longitudinal control of the vehicle. Finally, each of these components are simultaneously applied in the simulation environment using the robot operating system (ROS). As a result, we have achieved successful self-driving simulation. The key takeaway is that the perception and control strategies proposed can be easily transitioned towards physical implementation, through the use of ROS. The overall conclusion is that the CARLA simulation environment is a reliable workbench to test innovative solutions that could become technology enablers for the autonomous vehicle industry.
Prescinotti Vivan, Gabriel, Goberville, Nick, Asher, Zachary, Brown, Nicolas, Rojas, Johan
AVSC Best Practice for Metrics and Methods for Assessing Safety Performance of Automated Driving Systems (ADS)AVSC000062021033/25/2021
This AVSC Best Practice for Metrics and Methods for Assessing Safety Performance of Automated Driving Systems (ADS) (AVSC00006202103) recommends a set of metrics that may be used to assess ADS safety performance of the dynamic driving task (DDT). These metrics and methods are principally designed to provide evidence of safety performance for a manufacturer’s decision to deploy (and monitor) fleet-operated/managed SAE level 4 and 5 ADS-dedicated vehicles (ride-hailing or product delivery). This document lays out a performance-based, technology-neutral approach for measuring and analyzing safety performance. It supports long-term, socially-important safety goals (like reducing crashes). ADS safety performance metrics in this document support system-level analyses, i.e. they are practical to implement for any system regardless of architecture. The best practice provides: Metrics to Support ADS Safety Recommended Safety Outcomes Recommended Predictive Safety Metrics Methods for Assessing DDT Performance Metrics The metrics and methods provided in this document are intended for use by the technical community (developers, manufacturers, testers, etc.) to aid in the safe development and deployment of ADS. They may also be useful to stakeholders who have interest in better understanding the safety posture of ADS deployments.
Automated Vehicle Safety Consortium
Eco-profiling of Bio-epoxies via Life cycle AssessmentSAE-PP-002312/3/2021
Epoxies, synthesized from bisphenol-A (BPA) and epichlorohydrin (ECH), are predominantly used as coatings, adhesives, and as matrix material in fiber-reinforced composites for body-in-white (BiW) applications in the automotive sector. However, given the production of conventional epoxies from non-renewable petroleum resource and toxicity of BPA, several initiatives have been undertaken by researchers to synthesize alternative epoxies from various bio-sources that are free of BPA and exhibit similar mechanical performance. As a result, such bio-sourced epoxies are almost immediately termed as “eco-friendly”, despite the lack of comprehensive evaluation of their ecological performance that takes into account enhanced natural resource usage and associated impacts accompanying such epoxies. Hence, this work aims at addressing this gap by evaluating the environmental impacts of such bio-sourced epoxies via cradle-to-gate life cycle assessment to determine the genuine credentials of their ecofriendliness. Epoxies synthesized from three different bio-sources – namely, bark extractives, lignin, and triglyceride – were chosen, to evaluate their ecological performance. ReCiPe midpoint and endpoint methods were used to evaluate these epoxies in accordance with ISO 14040 and 14044 standards. Among the three bio-epoxies, lignin-based epoxy exhibits poor eco-performance mainly due to the use of large amount of chemicals during extraction of lignin, involving delignification and valorization. On the contrary, bio-epoxy synthesized from triglycerides were found to be eco-friendly compared to other bio-epoxies. All bio-epoxies are observed to contribute significantly to toxicity-related categories, mainly due to higher electricity consumption during both epoxy synthesis and manufacturing processes. Overall, this work sheds light on various avenues for synthesizing truly sustainable epoxies that exhibit mechanical performance comparable to their conventional counterparts.
Anthony, Lindsay, Jackson, Alyssa
2.0.104 - Tackling Three Critical Issues of Transportation: Environment, Safety and Congestion Via Semi-autonomous PlatooningSAE-PP-002142/1/2021
In recent years, platooning emerged as a realistic configuration for semi-autonomous driving. In the SARTRE project, simulation and physical tests were performed to validate the platooning system not only in testing facilities but also in conventional highways. Five vehicles were adapted with autonomous driving systems to have platooning functionalities, enabling to perform platoon tests and assess the feasibility, safety and benefits. Although the tested system was in a prototype, it demonstrated sturdiness and good functionality, allowing performing conventional road tests. First of all the fuel consumption decreased up to 16% in some configurations and different gaps between the vehicles were tested in order to establish the most suitable for platooning in terms of safety and economy. Additionally, the platooning technology enables a new level of safety in highways. Around 85% of the accident causation is the human factor. With platooning, the human factor is reduced almost entirely, making it a more efficient and safer system. With platooning, the traffic flow is more homogeneous with several vehicles travelling at the same speed and therefore accident situations are less probable while helping to maintain a steady traffic flow with no stop and go situations. For the same reason, the traffic flow is more manageable enabling advanced traffic management. Thus, the traffic congestion can be reduced and, again, reducing another source of emissions. All the potential advantages of a platooning technology which has been developed and tested are studied in this paper through test results and simulations.
Mutagaana, Festo
AVSC Best Practice for First Responder Interactions with Fleet-Managed Automated Driving System-Dedicated Vehicles (ADS-DVs)AVSC0000520201212/10/2020
This Automated Vehicle Safety Consortium (AVSC) Best Practice for First Responder Interactions with Fleet-Managed Automated Driving System-Dedicated Vehicles (ADS-DVs) (AVSC00005202012) recommends a common approach for describing the interactions and associated protocols for incorporating vital information, regarding first responder interactions with SAE level 4 and level 5 fleet operated vehicles, into on-going development and deployment plans. This best practice provides the following: Definitions of roles associated with emergency situations / processes, Descriptions of expected interactions (use cases) between first responders and ADS-DVs, Recommendations to address first responder and ADS-DV interactions, and Recommends a First Responder Interaction Plan framework. This best practice is also intended to provide the means for continued engagement and dialogue between first responders, developers, manufacturers, fleet operators and local communities to promote safe operations and consumer acceptance of fleet-managed ADS-DVs.
Automated Vehicle Safety Consortium
Safety-Relevant Guidance for On-Road Testing of Prototype Automated Driving System (ADS)-Operated VehiclesJ3018_202012 (Current)12/4/2020
This document provides preliminary1 safety-relevant guidance for in-vehicle fallback test driver training and for on-road testing of vehicles being operated by prototype conditional, high, and full (Levels 3 to 5) ADS, as defined by SAE J3016. It does not include guidance for evaluating the performance of post-production ADS-equipped vehicles. Moreover, this guidance only addresses testing of ADS-operated vehicles as overseen by in-vehicle fallback test drivers (IFTD). These guidelines do not address: Remote driving, including remote fallback test driving of prototype ADS-operated test vehicles in driverless operation. (Note: The term “remote fallback test driver” is included as a defined term herein and is intended to be addressed in a future iteration of this document. However, at this time, too little is published or known about this type of testing to provide even preliminary guidance.) Testing of driver support features (i.e., Levels 1 and 2), which rely on a human driver to perform part of the dynamic driving task (DDT) and to supervise the driving automation feature’s performance in real time. (Refer to SAE J3016.) Closed-course testing. Simulation testing (except for training purposes). Component-level testing. These guidelines also do not address prototype vehicle and IFTD performance data collection and retention. The collection of data invokes various legal and risk management considerations that users of this document should nevertheless bear in mind, such as: Maintaining auditable procedures and documentation. Adhering to applicable privacy laws and principles. Ensuring adequate data collection and recording integrity to support post-crash forensic analysis.
On-Road Automated Driving (ORAD) committee
Concept for an Aeronautical Design Standard (ADS) - Performance Specification for Autonomy Requirements of Military Air SystemsF-0076-2020-1631810/5/2020
This concept for an Aeronautical Design Standard for autonomous systems is methodology to determine the acceptable level of supervision for autonomy in military systems. The level of autonomy is directly related to the level of supervision required for the trust in the autonomous system functionality. The approach defines representative Autonomous Task Elements (ATEs), operational considerations, and levels of autonomy. Each ATE is characterized by an objective, description, and performance standards. Performance standards are expressed as metrics related to the trust in the autonomous behavior and the system’s capability to conduct the ATE. The capability of the autonomous system to meet performance standards is expressed as risk. This risk is compared to specified operational / allowable limits. The corresponding acceptable level of autonomy for the ATE is then determined using the probability of failure to meet the limits. The approach for this concept autonomy ADS parallels ADS 33E-PRF for the assignment of Levels of Handling Qualities based on measurable flight performance characteristics.
Preston, John
A Practical Fail-Operational Steering Concept02-13-03-001310/2/2020
Automated vehicles require some level of subsystem redundancy, whether to allow a transition time for driver re-engagement (L3) or continued operation in a faulted state (L4+). Highly automated vehicle developers need to have safe miles accumulated by vehicles to assess system maturity and experience new environments. This article presents a conceptual framework suggesting that hardware newly available to commercial vehicle application can be used to form a steering system that will remain operational upon a failure. The key points of a provisional safety case are presented, giving hope that a complete safety case is possible. This article will provide autonomous vehicle developers a view of a near term possibility for a highly automated commercial vehicle steering solution.
Pandy, Ananda, Pathuri, Nagamani, Salunke, Pranav, Subba, Srujana Sree, Williams, Dan
AVSC Best Practice for Data Collection for Automated Driving System-Dedicated Vehicles (ADS-DVs) to Support Event AnalysisAVSC000042020099/23/2020
As technology and functionality of vehicle systems change, so do data recording needs. In ADS-dedicated vehicles (DV), the ADS perceives the environment and handles vehicle motion control, i.e., the dynamic driving task (DDT), as described in SAE J3016. When an ADS takes the place of a human driver, its sensing, processing, and control systems necessitate new considerations for data recording. Data recording is important to crash reconstruction, system performance investigations, and event analysis. It enables industry-wide improvements in ADS safety. This best practice makes recommendations for the ADS-DV data needed to support: (1) information about what the ADS "saw" and "did" and (2) identify the technology-relevant factors that contributed to the event. The Best Practice for Data Collection for Automated Driving System Dedicated Vehicles (ADS-DVs) to Support Event Analysis makes recommendations for the collection, storage, and retrievability of onboard motor vehicle ADS event data. A list of 39 data elements are recommended for ADS-DVs to collect, including 14 new elements specific to ADS-DVs and 25 elements either previously defined or adapted from legacy standards in data collection.
Automated Vehicle Safety Consortium
AVSC Best Practice for Passenger-Initiated Emergency Trip InterruptionAVSC000032020066/30/2020
As passengers take rides in fleet-managed automated driving system-dedicated vehicles (ADS-DVs), they may feel the need to interrupt the trip due to a perceived emergency. There is currently no industry consensus on the proper balance between ADS passenger agency and the potential for introducing unexpected outcomes in dynamic traffic environments. In order to build public trust in automated vehicles, passengers should be given an option to exercise some type of control (agency) to intervene during situations they perceive as emergencies. Passenger-initiated emergency trip interruption features — however they manifest in a given vehicle - can help establish this confidence in ADS technologies. AVSC Best Practice for Passenger-Initiated Emergency Trip Interruption recommends processes surrounding aspects of passenger-initiated features in SAE level 4 and 5 fleet-managed ADS-DVs. It recommends criteria and processes for passenger initiation of these features from inside the vehicle; communication with passengers and fleet operations; enhanced diagnoses of the situation, interaction outside the vehicle with other road users, and general post-stop actions. Also, precautions against some types of foreseeable misuse are addressed. These recommendations apply to commercially available, deployed ADS-DV’s providing trips to people. The AVSC recommends that every fleet-managed SAE level 4 and 5 ADS-DV be equipped with a (PES) or (PEC) or both.
Automated Vehicle Safety Consortium
Taxonomy and Definitions for Terms Related to Cooperative Driving Automation for On-Road Motor VehiclesJ3216_202005 (Historical)5/7/2020
This document describes machine-to-machine (M2M) communication to enable cooperation between two or more participating entities or communication devices possessed or controlled by those entities. The cooperation supports or enables performance of the dynamic driving task (DDT) for a subject vehicle with driving automation feature(s) engaged. Other participants may include other vehicles with driving automation feature(s) engaged, shared road users (e.g., drivers of manually operated vehicles or pedestrians or cyclists carrying personal devices), or road operators (e.g., those who maintain or operate traffic signals or workzones). Cooperative driving automation (CDA) aims to improve the safety and flow of traffic and/or facilitate road operations by supporting the movement of multiple vehicles in proximity to one another. This is accomplished, for example, by sharing information that can be used to influence (directly or indirectly) DDT performance by one or more nearby road users. Vehicles and infrastructure elements engaged in cooperative automation may share information, such as state (e.g., vehicle position, signal phase), intent (e.g., planned vehicle trajectory, signal timing), or seek agreement on a plan (e.g., coordinated merge). Cooperation among multiple participants and perspectives in traffic can improve safety, mobility, situational awareness, and operations. However, nothing in this document is intended to suggest that driving automation requires such cooperation in order to be performed safely. Cooperative strategies may be enabled by the sharing of information in a way that meets the needs of a given application. The needs may be expressed in terms of performance characteristics, such as latency, transmission mode (e.g., one-way, two-way), range, privacy and security, and information content and quality. There are several potential technologies for communicating information between the subject vehicle and other participants. This document focuses on application-oriented functionality and does not imply the need for or require any specific functionality associated with communications protocols or the open systems interconnection model layers in a protocol stack. This document addresses the operational and tactical timescales of dynamic driving on ADS-operated vehicles, and excludes strategic functions such as trip scheduling and selection of destinations and waypoints. This information report is intended to facilitate communication and awareness for the design and anticipated development and validation of cooperative driving automation.
Cooperative Driving Automation(CDA) Committee
Chromaticity Requirements for Ground Vehicle Lamps and Lighting EquipmentJ578_202004 (Current)4/29/2020
This SAE Standard defines and provides a means for the control of colors employed in motor vehicle external lighting equipment, including lamps and reflex reflectors. The document applies to the overall effective color of light emitted by the device in any given direction, and not to the color of the light from a small area of the lens. It does not apply to pilot, indicator, or tell-tale lights.
Lighting Standard Practices Committee
AVSC Best Practice for Describing an Operational Design Domain: Conceptual Framework and LexiconAVSC000022020044/15/2020
An ADS-operated vehicle’s operational design domain (ODD) is defined by the manufacturer based on numerous factors. Research is underway at other organizations to define and organize ODD elements into taxonomies and other relational constructs. In order to enhance collaboration and communication between manufacturers and developers and transportation authorities, common terms and consistent frameworks are needed. The conceptual framework presented by Automated Vehicle Safety Consortium establishes a lexicon that can be used consistently by ADS developers and manufacturers responsible for defining their ADS ODD. A common framework and lexicon will reduce confusion, align expectations, and therefore build public trust, acceptance, and confidence. The guidance in this document is intended for: The technical community (e.g. manufacturers and developers) Public agencies (e.g. regulatory authorities) Infrastructure owner-operators The public This document, Best Practice for Describing an Operational Design Domain: Conceptual Framework and Lexicon is a critical first step. It offers a conceptual framework for manufacturers and developers to use when communicating with public agencies and the general public about their ADS’s ODD. It also details a list of potential variables with definitions that manufacturers and developers might use to describe certain aspects of the ODDs of their ADS-operated vehicles. It was developed with fleet-managed, SAE Level 4 vehicles in mind — i.e. vehicles requiring no human intervention to operate within their ODD. These vehicles are NOT privately owned.
Automated Vehicle Safety Consortium
Scenario Analysis and Control Comparison for a High Speed Autonomous Vehicle2020-01-07104/14/2020
This paper studies the simulation and control of an autonomous dragster. Four scenarios are provided that are critical to vehicle and driver safety in drag racing. Equations are then created to model the behavior during these safety scenarios. The use of a kinematic bicycle model and a Newtonian wheel stand model are discussed for plane-of-motion and out-of-plane vehicle movement, respectively. A separate controller is designed for each model by comparing different control methods. Proportional-Integral-Derivative (PID) control, optimal control, and model predictive control (MPC) are presented and applied to the models. The models are simulated from a speed of 75 m/s, being the estimated top speed of the research vehicle, up to a top speed of 150.5 m/s which is in alignment with the highest recorded speed of a dragster. The comparison of the control techniques yields MPC as superior for the bicycle model and PID as sufficient for the wheel stand model. Latency of the system is also discussed and accounted for in the system.
Donikian, Vatche, Bell, Joseph, Washington, Gregory
Modes of Automated Driving System Scenario Testing: Experience Report and Recommendations2020-01-12044/14/2020
With the widespread development of automated driving systems (ADS), it is imperative that standardized testing methodologies be developed to assure safety and functionality. Scenario testing evaluates the behavior of an ADS-equipped subject vehicle (SV) in predefined driving scenarios. This paper compares four modes of performing such tests: closed-course testing with real actors, closed-course testing with surrogate actors, simulation testing, and closed-course testing with mixed reality. In a collaboration between the Waterloo Intelligent Systems Engineering (WISE) Lab and AAA, six automated driving scenario tests were executed on a closed course, in simulation, and in mixed reality. These tests involved the University of Waterloo’s automated vehicle, dubbed the “UW Moose”, as the SV, as well as pedestrians, other vehicles, and road debris. Drawing on both data and the experience gained from executing these test scenarios, the paper reports on the advantages and disadvantages of the four scenario testing modes, and compares them using eight criteria. It also identifies several possible implementations of mixed-reality scenario testing, including different strategies for data mixing. The paper closes with twelve recommendations for choosing among the four modes.
Antkiewicz, Michał, Kahn, Maximilian, Ala, Michael, Czarnecki, Krzysztof, Wells, Paul, Acharya, Atul, Beiker, Sven
Nonlinear Model Predictive Control of Autonomous Vehicles Considering Dynamic Stability Constraints2020-01-14004/14/2020
Autonomous vehicle performance is increasingly highlighted in many highway driving scenarios, which leads to more priorities to vehicle stability as well as tracking accuracy. In this paper, a nonlinear model predictive controller for autonomous vehicle trajectory tracking is designed and verified through a real-time simulation bench of a virtual test track. The dynamic stability constraints of nonlinear model predictive control (NLMPC) are obtained by a novel quadrilateral stability region criterion instead of the conventional phase plane method using the double-line region. First, a typical lane change scene of overtaking is selected and a new composited trajectory model is proposed as a reference path that combines smoothness of sine wave and comfort of linear functional path. Reference lateral velocity, azimuth angle, yaw rate, and front wheel steering angle are subsequently taken into account. Then, by establishing a nonlinear vehicle dynamics model where Magic Formula of nonlinear tire model is adapted, the quadrilateral vehicle stability region is defined in consideration of designed velocity, road adhesion coefficient, and front wheel steering angle. Working condition-variant constraints determined by the boundaries of the quadrilateral region are subsequently obtained to guarantee the stability and vehicle performance. Finally, a nonlinear motion state space model with measured and unmeasured disturbance for NLMPC tracking maneuver is proposed, Meanwhile, a multi-objective cost function based on track error, ride comfort, and the smoothness of control derivative is established. Laguerre functions are applied to design optimal control trajectory and Hildreth’s quadratic programming procedure is introduced to find converged solutions meeting constraints derived from previously investigated quadrilateral stability region for sake of lightening computation load and finding better numerically conditioned solutions of control when NLMPC is implemented online. The configuration of a real-time virtual test track is explained and the NLMPC algorithm is validated. The simulation and experiment results are illustrated to show the effectiveness of the designed nonlinear model predictive control scheme under the test of the overtaking scene compared with the conventional driver control. This work may provide a useful basis for researches of autonomous vehicle lane change in terms of track accuracy, ride comfort as well as stability.
Chen, Xunjie, Wu, Guangqiang, Ren, Meng
Control Model of Automated Driving Systems Based on SOTIF Evaluation2020-01-12144/14/2020
In partially automated and conditionally automated vehicles, a part of the work of human drivers is replaced by the system, and the main source of safety risks is no longer system failures, but non-failure risks caused by insufficient system function design. The absence of unreasonable risk due to hazards resulting from functional insufficiencies of the intended functionality or by reasonably foreseeable misuse by persons, is referred to as the Safety Of The Intended Functionality. Drivers have the responsibility to supervise the automated driving system. When they don't agree with the operation behavior of the system, they will interfere with the instructions. However, this may lead to potential risks. In order to discover the causes of human misuse, this paper takes the trust feeling between the driver and the automated driving system as the starting point, and based on the collected data of track test, establishes the evaluation indicator -- degree of confidence to show the trust feeling between the driver and the automated system. Degree of confidence is a comprehensive interpretation of the driver's physical and psychological feelings. In the process of track test, we simultaneously collect the dynamics indicators of the vehicle. After the test, the drivers' driving feeling was evaluated by questionnaire. Then, the relationship between objective indicator and subjective score was established by machine learning method, and the development of evaluation indicator was completed. Finally, this paper optimizes the automatic driving motion planning algorithm based on this indicator, and verifies the effectiveness of the algorithm through simulation.
Guo, Mengge, Shang, Shiliang, Haifeng, Cui, Zhang, Kaijiong, Deng, Weishun, Zhang, Xi, Yu, Fan
Noise, Vibration, and Harshness Considerations for Autonomous Vehicle Perception Equipment2020-01-04824/14/2020
Automakers looking to remake their traditional vehicle line-up into autonomous vehicles, Noise, Vibration, and Harshness (NVH) considerations for autonomous vehicles are soon to follow. While traditional NVH considerations still must be applied to carry-over systems, additional components are required for an autonomous vehicle to operate. These additional components needed for autonomy also require NVH analysis and optimization. Autonomous vehicles rely on a suite of sensors, including Light Detection and Ranging (LiDAR) and cameras placed at optimal points on the vehicle for maximum coverage and utilization. In this study, the NVH considerations of autonomous vehicles are examined, focusing on the additional perception equipment installed in autonomous vehicles. In particular, the nature of modifications to existing vehicles to increase the level of autonomy, and the associated NVH characteristics of these alterations, are reviewed with suggestions for future application to autonomous vehicles. A case study in the design of an original autonomous vehicle based on a production all-electric car, a 2017 Chevrolet Bolt, is outlined. A detailed description of the NVH design and verification process for this vehicle is provided, with results giving insight into the NVH design of autonomous vehicles and the challenges that are created.
Gates, Charlie, Bastiaan, Jennifer, Jadhav, Prashant, Baqersad, Javad, Peters, Diane
On Perception Safety Requirements and Multi Sensor Systems for Automated Driving Systems2020-01-01014/14/2020
One major challenge in designing SAE level 3-5 Automated Driving Systems (ADS) is to define requirements for the perception system that would enable argumentation for safe operation. The safety requirements on the perception system can only be fulfilled through redundancy in the sensor hardware. It is, however, a challenge to specify the redundancy that is required in the sensor system. Safe operation for an ADS is significantly more difficult compared to advanced driver assistance systems (ADAS). The safety argumentation for ADAS typically argues that in case of a failure in the sensor array a fail-silent behavior is acceptable because the human driver can take control of the vehicle back. This argumentation however is not possible when developing level 4 or higher automation. This paper investigates prerequisites for applying a systematic methodology for analyzing redundancy in a multi-sensor system and the relation to a conceptual ADS functional architecture. This analysis must address the complexity that comes with partially overlapping sensor data from different sensors and considers variations in performance and characteristics due to changes in the environmental conditions. The paper introduces the term incomplete redundancy and presents a systematic methodology for analyzing redundancy. The aim is to provide arguments for how several sensors in a system, when appropriately combined, meet an assigned safety requirement on a higher level. Each sensor will then be assigned a certain responsibility and contributes with a sub-set of information. A set of questions of importance to address as a foundation for such a methodology are defined and discussed. The definitions of redundancy and independence between sensors are discussed as well as contract-based functional safety to adapt to different environmental and operating conditions.
Cassel, Anders, Bergenhem, Carl, Christensen, Ole Martin, Heyn, Hans-Martin, Leadersson-Olsson, Susanna, Majdandzic, Mario, Sun, Peng, Thorsén, Anders, Trygvesson, Jörgen
Analysis of LiDAR and Camera Data in Real-World Weather Conditions for Autonomous Vehicle Operations2020-01-00934/14/2020
Autonomous vehicle technology has the potential to improve the safety, efficiency, and cost of our current transportation system by removing human error. With sensors available today, it is possible for the development of these vehicles, however, there are still issues with autonomous vehicle operations in adverse weather conditions (e.g. snow-covered roads, heavy rain, fog, etc.) due to the degradation of sensor data quality and insufficiently robust software algorithms. Since autonomous vehicles rely entirely on sensor data to perceive their surrounding environment, this becomes a significant issue in the performance of the autonomous system. The purpose of this study is to collect sensor data under various weather conditions to understand the effects of weather on sensor data. The sensors used in this study were one camera and one LiDAR. These sensors were connected to an NVIDIA Drive Px2 which operated in a 2019 Kia Niro. Two custom scenarios (static and dynamic objects) were chosen to collect sensor data operating in four real-world weather conditions: fair, cloudy, rainy, and light snow. An algorithm developed herein was used to provide a method of quantifying the data for comparison against the other weather conditions. The results from these performance algorithms show that sensor data quality degrades by an average of 13.88% for static objects and 16.16% for dynamic objects while operating in these conditions, with operations in rain proving to have the most significant effect on sensor data degradation. From this study, it is hypothesized that advancements in data processing algorithms can improve the usability of this degraded data. In future work, we seek to explore fault-tolerant sensor fusion algorithms that can overcome the effects of adverse weather.
Goberville, Nick, El-Yabroudi, Mohammad, Omwanas, Mark, Rojas, Johan, Meyer, Rick, Asher, Zachary, Abdel-Qader, Ikhlas
Observer for Faulty Perception Correction in Autonomous Vehicles2020-01-06944/14/2020
Operation of an autonomous vehicle (AV) carries risk if it acts on inaccurate information about itself or the environment. The perception system is responsible for interpreting the world and providing the results to the path planning and other decision systems. The perception system performance is a result of the operating state of the sensors, e.g. is a sensor in fault or being adversely affected by the weather or environmental conditions, and approach to sensor measurement interpretation. We propose a trailing horizon switched system observer that minimizes the difference between reference tracking values developed from sensor fusion performed at an upper level and the values from a potentially faulty sensor based upon the convex combination of different sensor observation model outputs; the sensor observations models are associated with different sensor operating errors. The preferred observer target is a stationary landmark so as to remove additional uncertainty resulting from tracking of moving targets. Results for five scenarios show the observer identifies the appropriate sensor model in no more than a few sample intervals.
Omwansa, Mark, Meyer, Rick, Asher, Zachary, Goberville, Nick
Autonomous Vehicles Camera Blinding Attack Detection Using Sequence Modelling and Predictive Analytics2020-01-07194/14/2020
Autonomous vehicles are waiting to address the global automotive mobility challenges through an intelligent smart transportation system, which includes advanced sensor-actuator configurations to control, navigate, and drive the vehicles. Multi-sensor data fusion from the key sensors such as camera, radar, and lidar is used to achieve the environmental perception for autonomous vehicles by capturing the various attributes of the environment. Cameras are the dominant sensors to achieve the perception by providing vision capability to vehicles. The direct interface of the cameras with the dynamic driving environment carries numerous attack surfaces on the camera. Blinding attacks on the cameras are one of the critical attacks with an intention to blind the cameras either fully or partially by projecting light into the cameras to hide the objects which results in failure in object detection. Here, the blinding attack detection approach is proposed which detects the blinding attacks on the camera in a dynamic driving environment by camera data predictive analytics. The proposed system predicts the future next frame of the video at each time instance and compares the received frame from the camera with the predicted frame at that instance to detect the blinding attacks. The incoming frames from the camera are sequentially modeled using a convolutional encoder-decoder neural network to predict the consecutive future frames, and the predicted frames are compared with the received camera frames to identify the similarity measure between the predicted and incoming camera frames of the same instance. Further, the approach detects the blinding on the camera, if the similarity measure calculated falls below a fixed threshold. The similarity measure which is inversely proportional to the amount of blinding is used to identify the blinding attacks. The predictive analytics of the sequentially modeled video frames with similarity measurement is used for the successful detection of blinding attacks.
D H, Sharath Yadav, Ansari, Asadullah
Continuous Integration as Mandatory Puzzle Piece for the Success of Autonomous Vehicles2020-01-00874/14/2020
The transition to autonomous driving technology is widely discussed topic today. In order to make autonomous vehicles work safely in the long run it will be a necessity to keep their software up to date at any time. The challenge is that software released with today’s traditional release methods for vehicle updates is not deployed fast enough. Newly discovered corner cases or glitches in the design could restrict the usage of entire fleets for long time. This paper discusses the use of continuous integration methods implemented into the automotive system development in order to keep up with the pace needed to make the new technology a success, and accepted by the users. The development process has to contain smart branching strategies for fast turn around. It is mandatory to have a frozen and stable branch to release hotfixes in case of need, a validation branch with feature lock in order to stabilize, and a feature branch heavy development space that is supported by full system regression testing from the very beginning. The change content for validation per test execution has to be limited to minimum in order to support fast issue identification and root cause analysis. A sophisticated end to end continuous integration and validation process applied on the highest system integration level can achieve turn around times measured in hours and not in weeks.
Rohde, Florian
Rain-Adaptive Intensity-Driven Object Detection for Autonomous Vehicles2020-01-00914/14/2020
Deep learning based approaches for object detection are heavily dependent on the nature of data used for training, especially for vehicles driving in cluttered urban environments. Consequently, the performance of Convolutional Neural Network (CNN) architectures designed and trained using data captured under clear weather and favorable conditions, could degrade rather significantly when tested under cloudy and rainy conditions. This naturally becomes a major safety issue for emerging autonomous vehicle platforms relying on CNN based object detection methods. Furthermore, despite a noticeable progress in the development of advanced visual deraining algorithms, they still have inherent limitations for improving the performance of state-of-the-art object detection. In this paper, we address this problem area by make the following contributions. We systematically study and quantify the influence of a wide range of rain intensities on the performance of popular deep learning based object detection that is trained with clear visual data. We show that even low rain intensities could significantly degrade the performance of object detection trained using clear visuals. Subsequently, we propose a Rain-Adaptive Intensity-Driven (RAID) deep learning framework for object detection under a variety of rain intensities. Controlled experiments based on rain simulations, which are seamlessly integrated with real visual data captured by moving vehicles in truly cluttered urban environments, show the superiority of the proposed RAID framework as compared with state-of-the-art deraining methods in conjunction with popular deep learning based object detection.
Hnewa, Mazin, Radha, Hayder
Understanding How Rain Affects Semantic Segmentation Algorithm Performance2020-01-00924/14/2020
Research interests in autonomous driving have increased significantly in recent years. Several methods are being suggested for performance optimization of autonomous vehicles. However, weather conditions such as rain, snow, and fog may hinder the performance of autonomous algorithms. It is therefore of great importance to study how the performance/efficiency of the underlying scene understanding algorithms vary with such adverse scenarios. Semantic segmentation is one of the most widely used scene-understanding techniques applied to autonomous driving. In this work, we study the performance degradation of several semantic segmentation algorithms caused by rain for off-road driving scenes. Given the limited availability of datasets for real-world off-road driving scenarios that include rain, we utilize two types of synthetic datasets. The first dataset is a pure synthetic rainy dataset which considers the rain droplets on a camera lens, which is suitable for an autonomous vehicle with outside-mounted cameras. This data is generated by the MAVS simulator. In the second dataset, we take good-weather imagery and artificially incorporate rain streaks. By investigating different simulated rain rates, we quantify the performance of such algorithms and witness the severe performance degradation with increasing rain density. We also propose and analyze two methods to obtain the robust performance of segmentation algorithms for both clear and rainy weather.
Sharma, Suvash, Goodin, Chris, Doude, Matthew, Hudson, Christopher, Carruth, Daniel, Tang, Bo, Ball, John
A Novel Velocity Planner for Autonomous Vehicle Considering Human Driver’s Habits2020-01-01334/14/2020
In automatic driving application, the velocity planner can be considered as a key factor to ensure the safety and comfort. One of the most important tasks of the velocity planner is to simulate the velocity characteristics of human drivers. In this paper, two Driver In-the-Loop (DIL) experiments are designed to explain velocity characteristics of human drivers. In the first experiment, static obstacles are placed on both sides of the straight road to shorten the cross range that vehicles can driver across. Moreover, different cross ranges are set to study the influence of the steering wheel error. In the second experiment, velocity characteristics are investigated under the condition of different road widths and curvatures in a U-turn road contour. In both tests, different drivers’ preview behavior is analyzed through the operation of throttle, braking, and steering. From the results we could see the change of vehicle speed depends largely on the traffic environment at the driver’s preview point. On this basis, a novel velocity planner is proposed. Firstly, a target velocity in preview terminal point is calculated. The calculation of the velocity is based on two indicators-the driver’s driving & operating ability, and the degree of visual restriction. The former refers to the ability of the driver to maintain the driveway as well as the control ability of the vehicle stability, and the latter is related to the uncertainty of the environment. Subsequently, the smooth velocity profiles that connect the initial point and the preview terminal point are generated based on the convex optimization. Finally, the simulation results show that this velocity planner possesses good human-like performance, considering the human-vehicle-road coordination. This study is useful to customize velocity planning for autonomous vehicle so as to improve the acceptability of the specific human driver.
Cui, Zongwei, Guo, Xuexun, Pei, Xiaofei
Platooning Vehicles Control for Balancing Coupling Maintenance and Trajectory Tracking - Feasibility Study Using Scale-Model Vehicles2020-01-01284/14/2020
Recently, car-sharing services using ultra-compact mobilities have been attracting attention as a means of transportation for one or two passengers in urban areas. A platooning system consisting of a manned leader vehicle and unmanned follower vehicles can reduce vehicle distributors. We have proposed a platooning system which controls vehicle motion based on the relative position and posture measured by non-contact coupling devices installed between vehicles. The feasibility of the coupling devices was validated through a HILS experiment. There are two basic requirements for realizing our platooning system; (1) all devices must remain coupled and (2) follower vehicles must be able to track the leader vehicle trajectory. Thus, this paper proposes two vehicle control method for satisfying those requirements. They are the “device coupling and trajectory tracking merging method” and the “trajectory shifting method”. The device coupling and trajectory tracking merging method consisting of a coupling keeping controller and a trajectory tracking controller. The predominant controller is chosen according to the amount of the coupling device error and the trajectory tracking error. The trajectory shifting method shifts the tracking target trajectory to keep the device coupled. The shifting amount is decided by the estimated turning radius of the leader vehicle. Platooning experiments using two 1:16 scale-model vehicles has been performed on the experiment course containing a straight section and a circular section. Experiment results revealed that the device coupling and trajectory tracking merging method can maintain the coupling of the device while limiting the trajectory tracking error to a certain range. Though the trajectory shifting method can reduce the coupling device error, it fails on both device coupling keeping and trajectory error limiting, owing to the inadequacy in estimating the turning radius of the leader.
Fukui, Rui, Ye, Qiwei, Suzuki, Ayumi, Warisawa, Shin’ichi
Effects of a Probability-Based Green Light Optimized Speed Advisory on Dilemma Zone Exposure2020-01-01164/14/2020
Green Light Optimized Speed Advisory (GLOSA) systems have the objective of providing a recommended speed to arrive at a traffic signal during the green phase of the cycle. GLOSA has been shown to decrease travel time, fuel consumption, and carbon emissions; simultaneously, it has been demonstrated to increase driver and passenger comfort. Few studies have been conducted using historical cycle-by-cycle phase probabilities to assess the performance of a speed advisory capable of recommending a speed for various traffic signal operating modes (fixed-time, semi-actuated, and fully-actuated). In this study, a GLOSA system based on phase probability is proposed. The probability is calculated prior to each trip from a previous week’s, same time-of-day (TOD) and day-of-week (DOW) period, traffic signal controller high-resolution event data. By utilizing this advisory method, real-time communications from the vehicle to infrastructure (V2I) become unnecessary, eliminating data-loss related issues. The effects of three different advice approaches (conservative, balanced, and aggressive) on dilemma zone exposure are analyzed. Proof of concept is carried out by simulating drives through a test-route composed of an arterial that had historical high-resolution traffic signal event logs for a series of actuated-coordinated traffic signals during different TOD and DOW. A comparison was performed between unadvised and GLOSA advised trips obtained from approximately 486,000 simulated trajectories. Results were obtained by analyzing the vehicle’s probability of stopping from utilizing Traffic Engineering dilemma zone theory. Reductions of 93% in the amount of hard brakings and 96% in the number of crossings through red light were observed with the proposed system. This data suggests the feasibility of a probability-based advisory, as well as the viability of utilizing the proposed GLOSA system to minimize dilemma zone exposure.
Saldivar-Carranza, Enrique, Li, Howell, Kim, Woosung, Mathew, Jijo, Bullock, Darcy, Sturdevant, James
Trajectory Planning and Tracking for Four-Wheel-Steering Autonomous Vehicle with V2V Communication2020-01-01144/14/2020
Lane-changing is a typical traffic scene effecting on road traffic with high request for reliability, robustness and driving comfort to improve the road safety and transportation efficiency. The development of connected autonomous vehicles with V2V communication provide more advanced control strategies to research of lane-changing. Meanwhile, four-wheel steering is an effective way to improve flexibility of vehicle. The front and rear wheels rotate in opposite direction to reduce the turning radius to improve the servo agility operation at the low speed while those rotate in same direction to reduce the probability of the slip accident to improve the stability at the high speed. Hence, this paper established Four-Wheel-Steering(4WS) vehicle dynamic model and quasi real lane-changing scenes to analyze the motion constraints of the vehicles. Then, the polynomial function was used for the lane-changing trajectory planning and the extended rectangular vehicle model was established to get vehicle collision avoidance condition. Vehicle comfort requirements and lane-changing efficiency were used as the optimization variables of optimization function and the control of trajectory tracking can be obtained by using model predictive control (MPC) method. A lane-changing model based on steering characteristics and safety distance with the system of V2V communication and collaboration strategy was established. The lane-changing trajectory was simulated by MATLAB and the results showed that the lane-changing trajectory can safely realize the lane-changing behavior of 4WS autonomous vehicles.
Ma, Fangwu, Shen, Yucheng, Nie, Jiahong, Li, Xiyu, Yang, Yu, Wang, Jiawei, Wu, Guanpu
Effect of Space Weather on Autonomous Vehicle Navigation2020-01-01404/14/2020
Autonomous vehicle systems integrate multiple information systems. Navigation is reliant on Global navigation satellite systems (GNSS) such as the global positioning systems (GPS) which are supported by a satellite network. However, satellites and radio signals are subject to interference from sunspots. Sunspots happen on regular cycles at varying strengths but their occurrence can’t be exactly predicted. The likelihood of a severe solar event is roughly twelve percent per decade; consequently, solar events are likely to impair navigation. Results will show the probability of each event and its impact on autonomous vehicle navigation. In the worst case scenario, satellites could even be permanently damaged by severe sunspots. As autonomous vehicles become a more significant portion of the economy, it is necessary that they have resilience to operate in extreme conditions. Alternative navigation procedures are proposed to enhance the resiliency of autonomous vehicles. Artificial intelligence related to place identification with relative geographic navigation and storage of most common routes is an option.
Hoback, Alan
Design of a Mild Hybrid Electric Vehicle with CAVs Capability for the MaaS Market2020-01-14374/14/2020
There is significant potential for connected and autonomous vehicles to impact vehicle efficiency, fuel economy, and emissions, especially for hybrid-electric vehicles. These improvements could have large-scale impact on oil consumption and air-quality if deployed in large Mobility-as-a-Service or ride-sharing fleets. As part of the US Department of Energy's current Advanced Vehicle Technology Competition (AVCT), EcoCAR: The Mobility Challenge, Mississippi State University’s EcoCAR Team is redesigning and doing the development work necessary to convert a conventional gasoline spark-ignited 2019 Chevy Blazer into a hybrid-electric vehicle with SAE Level 2 autonomy. The target consumer segments for this effort are the Mobility-as-a-Service fleet owners, operators and riders. To accomplish this conversion, the MSU team is implementing a P4 mild hybridization strategy that is expected to result in a 30% increase in fuel economy over the stock Blazer. MATLAB models of the vehicle system shows the potential for additional improvement with the use of connected and autonomous features in the vehicle. This paper presents the design rationale for selection of the P4 strategy, vehicle modeling, and fuel economy simulation results completed during Year 1 of the competition. A detailed discussion of further improvements arising from incorporating connected and autonomous technology strategies, focusing on longitudinal control methods is also presented.
Taoudi, Amine, Haque, Moinul Shahidul, Strzelec, Andrea, Follett, Randolph
Development of a Procedure to Correlate, Validate and Confirm Radar Characteristics of Surrogate Targets for ADAS Testing2020-01-07164/14/2020
Surrogate targets are used throughout the automotive industry to safely and repeatably test Advanced Driver Assistance Systems (ADAS) and will likely find similar applications in tests of Automated Driving Systems. For those test results to be applicable to real-world scenarios, the surrogate targets must be representative of the real-world objects that they emulate. Early target development efforts were generally divided into those that relied on sophisticated radar measurement facilities and those that relied on ad-hoc measurements using automotive grade equipment. This situation made communication and interpretation of results between research groups, target developers and target users difficult. SAE J3122, “Test Target Correlation - Radar Characteristics”, was developed by the SAE Active Safety Systems Standards Committee to address this and other challenges associated with target development and use. J3122 addresses four topics. First, it describes standardized equipment and procedures for making various types of calibrated radar measurements using automotive grade equipment, with minimal measurement site restrictions. Second, a correlation procedure is provided that is used to define validity regions and properties of representative real-world objects. Third, a validation procedure is provided for comparing candidate targets against measurements of representative objects using an objective correlation score. Finally, a confirmation procedure is provided for checking in-use targets to verify that they continue to be acceptable for testing. This paper describes each of these topics as well as the process development.
Silberling, Jordan, Nicols, George, Buller, William, Lenkeit, John
Runtime Active Safety Risk-Assessment of Highly Autonomous Vehicles for Safe Nominal Behavior2020-01-01074/14/2020
Fatal crashes involving automated driving systems, has been raising the concern of minimum standard requirement for safety, reliability and performance required for Autonomous Driving System (ADS)/Advanced Driver Assistance System (ADAS) before this cutting-edge technology takes on public roads. Hence, in order to ensure necessary safety requirements of ADS/ADAS systems we propose a runtime active safety assurance module known as SConSert. SConSert performs dynamic risk assessment of “Sensing, Planning and Action module of ADS/ADAS”; to provide minimal risk maneuver in any given driving scenario. The dynamic risk assessment of ADS/ADAS system is based on the operational design domain (ODD) knowledge of the driving scenario plus the sensor capability, ADS/ADAS algorithm requirement and capability, and finally smooth and collision free maneuver requirement. So, the main concept behind SConSert is runtime derivation of situational and conditional set of contracts for a given driving scenario and ADS/ADAS system ODD; fulfillment or violation of which can help in runtime dynamic risk assessment of ADS/ADAS to plan minimal safe behavior such that necessary safety requirements can be achieved. Finally, through experiment we show that proposed runtime active assurance safety module can handle complex driving scenario, and present simulation and experimental results that emphasizes the importance of the proposed runtime safety assurance module and shows that the proposed system is capable of performing runtime dynamic risk assessment in order to keep the automated driving systems always within the safe sate that is the automated driving system always perform within its ODD.
Rathour, Swarn Singh, Ishigooka, Tasuku, Otsuka, Satoshi, MARTIN, RAUL
Cooperative Mandatory Lane Change for Connected Vehicles on Signalized Intersection Roads2020-01-08894/14/2020
This paper presents a hierarchical control architecture to coordinate a group of connected vehicles on signalized intersection roads, where vehicles are allowed to change lane to follow a prescribed path. The proposed hierarchical control strategy consists of two control levels: a high level controller at the intersection and a decentralized low level controller in each car. In the hierarchical control architecture, the centralized intersection controller estimates the target velocity for each approaching connected vehicle to avoid red light stop based on the signal phase and timing (SPAT) information. Each connected vehicle as a decentralized controller utilizes model predictive control (MPC) to track the target velocity in a fuel efficient manner. The main objective in this paper is to consider mandatory lane changes. As in the realistic scenarios, vehicles are not required to drive in single lane. More specifically, they more likely change their lanes prior to signals. Hence, the vehicle decentralized controllers must prepare to cooperate with the vehicle that has a mandatory lane change request (host vehicle). The cooperative mandatory lane change is accomplished by inserting a virtual vehicle on the host vehicle’s target lane. The simulation results show the advantage of our proposed approach on both the lane change duration and vehicle fuel economy.
Du, Zhiyuan, Xu, Bin, Pisu, Pierluigi
Prevention of Snow Accretion on Camera Lenses of Autonomous Vehicles2020-01-01054/14/2020
With the rapid development of artificial intelligence, the autonomous vehicles (AV) have attracted considerable attention in the automotive industry. However, different factors negatively impact the adoption of the AVs, delaying their successful commercialization. Accretion of atmospheric icing, especially wet snow, on AV sensors causes blockage on their lenses, making them prone to lose their sight, in turn, increasing potential chances of accidents. In this study, two different designs are proposed in order to prevent snow accretion on the lenses of AVs via air flow across the lens surface. In both designs, lenses made of plain glass and superhydrophobic coated glass surfaces are tested. While some researchers have shown promise of water repellency on superhydrophobic surfaces, more snow accretion is observed on the superhydrophobic surfaces, when compared to the plain glass lenses. In the experiments, snow is formed using a novel snow gun inside a walk-in cold room connected to a wind tunnel that can reach wind speeds of up to 40 mph. It is observed that the air flow over the lens significantly reduces the accretion of snow on the lens and could maintain the lens clean for up to 20 mph wind velocities. However, at LWC values of approximately 28%, the stickiness of the snowflakes increases, enhancing snow accretion on the lenses and translating to significant loss of sight. The high stickiness of the snowflakes along with high wind speeds leads to increased blockage of the AV lenses.
Mohammadian, Behrouz, Sarayloo, Mehdi, Heil, Jamie, Sojoudi, Hossein, Robertson, Michael, Hong, Haiping, Tran, Tommy, Patil, Sunil, Krishnan, Venkatesh
Robust Sensor Fused Object Detection Using Convolutional Neural Networks for Autonomous Vehicles2020-01-01004/14/2020
Environmental perception is considered an essential module for autonomous driving and Advanced Driver Assistance System (ADAS). Recently, deep Convolutional Neural Networks (CNNs) have become the State-of-the-Art with many different architectures in various object detection problems. However, performances of existing CNNs have been dropping when detecting small objects at a large distance. To deploy any environmental perception system in real world applications, it is important that the system achieves high accuracy regardless of the size of the object, distance, and weather conditions. In this paper, a robust sensor fused object detection system is proposed by utilizing the advantages of both vision and automotive radar sensors. The proposed system consists of three major components: 1) the Coordinate Conversion module, 2) Multi level-Sensor Fusion Detection (MSFD) system, and 3) Temporal Correlation filtering module. The proposed MSFD system employs the principles of artificial intelligence beyond simple comparison of data variance of the sensors. And then, its performance is further improved by using the temporal correlation information with an adaptive threshold scheme. The proposed system is evaluated with the collected video data (6,854 image frames with 18,918 labeled objects). Based on the laboratory testing and in-vehicle validation, the proposed system demonstrates its high accuracy for detecting any size of objects in real-world data.
Park, Jungme, Jayachandran Raguraman, Sriram, Aslam, Aakif, Gotadki, Shruti
Automated Driving System Data LoggerJ3197_202004 (Historical)4/11/2020
This recommended practice provides common data output formats and definitions for a variety of data elements that may be useful for analyzing the performance of automated driving system (ADS) during an event that meets the trigger threshold criteria specified in this document. The document is intended to govern data element definitions, to provide a minimum data element set, and to specify a common ADS data logger record format as applicable for motor vehicle applications. The data elements defined in this document are unique to Levels 3, 4, or 5 ADS features, as defined by SAE J3016, and provide additional background of the events leading up to a crash or crash-like event. The data from sensors such as camera(s), LiDAR(s) etc. will provide information in the absence of a human driver. The data included in the ADS data logger is expected to be used in conjunction with the SAE J1698 EDR record and traditional accident reconstruction analysis. The event data recorder (EDR) and ADS data logger will capture information leading up to the triggered event, at a minimum. ADS technology is still being developed and is not yet commercially deployed. Therefore, this SAE Recommended Practice is intended as a guide toward standard practice and is subject to change to keep pace with experience and technical advances.
Event Data Recorder Committee
Unsettled Topics Concerning Automated Driving Systems and the Development Ecosystem SAAS Demo 10-17-23EPR20200043/17/2020
With over 100 years of operation, the current automobile industry has settled into an equilibrium with the development of methodologies, regulations, and processes for improving safety. In addition, a nearly $2-trillion market operates in the automotive ecosystem with connections into fields ranging from insurance to advertising. Enabling this ecosystem is a well-honed, tiered supply chain and an established development environment. Autonomous vehicle (AV) technology is a leap forward for the existing automotive industry; now the automobile is expected to manage perception and decision-making tasks. The safety technologies associated with these tasks were presented in an earlier SAE EDGE™ Research Report, “Unsettled Technology Areas in Autonomous Vehicle Test and Validation.” In a later SAE EDGE™ Research Report, “Unsettled Topics Concerning Automated Driving Systems and the Transportation Ecosystem,” senior executives from the automotive ecosystem explored the impact of AV technology as they faced the prospect of this disruptive technology entering their marketplace. Interestingly, stable use-models and market penetration were all gated primarily by the demonstration of AV safety. Building on these previous verification and validation (V&V)-related reports, “Unsettled Topics Concerning Automated Driving Systems and the Development Ecosystem” explores the open issues in the shift of the development and supplier environment toward a new AV-enabled future. NOTE: SAE EDGE™ Research Reports are intended to identify and illuminate key issues in emerging, but still unsettled, technologies of interest to the mobility industry. The goal of SAE EDGE™ Research Reports is to stimulate discussion and work in the hope of promoting and speeding resolution of identified issues. SAE EDGE™ Research Reports are not intended to resolve the challenges they identify or close any topic to further scrutiny.
Razdan, Rahul
GM deploying electricals of the future20AUTP03_083/1/2020
In tandem with its corporate strategy to “skip” vehicle hybridization and focus solely on electric vehicles (EVs) for future propulsion (see cover story), General Motors recently offered media a deeper look at the company's all-new vehicle electrical architecture that was first confirmed in mid-2019, saying the vehicle-wide electrical system is associated with more than 100 new patents. The advanced “digital vehicle platform,” known internally as Global B, debuted with the 2020 Cadillac CT5 sedan. It serves as the electrical foundation for the 2020 Chevrolet Corvette Stingray and the 2021 Chevrolet Tahoe and Suburban, GM Yukon and Cadillac Escalade and will carry into all of the company's future conventionally-powered and EV models.
Visnic, Bill
Items per page:
1 – 50 of 302