Development Of Autonomous Vehicles Research Articles

Efficient Clustering of Visible Light Communications in VANET

The deployment of vehicular ad hoc network (VANET) is crucial to the development of autonomous vehicles. Radio frequency (RF) technology has been employed to transmit messages between vehicles and infrastructure in VANET. However, the limited RF bands may cause interference when vehicles transmit messages in a high-density environment. Moreover, when numerous vehicles transmit messages to the infrastructure at the same time, the simultaneous transmissions may cause channel congestion. While the issue of signal interference can be solved by the techniques of Visible Light Communication (VLC), vehicle clustering can be employed to improve the transmission performance of vehicles. VLC is an emerging technology with the advantage of immunity to electromagnetic interference. The technique of vehicle clustering categorizes vehicles into different sets, where each set has a leader for intra-cluster messaging. In this work, we present a clustering algorithm for VANET based on VLC. Our algorithm estimates the positions of vehicles based on their current movements. Then, it selects cluster heads based on the number of neighboring vehicles and generates clusters. We evaluate the performance of our scheme for both urban and highway scenarios. The simulation results show that the proposed algorithm can minimize the number of clusters and improve the transmission data rate for vehicles.

Autonomous Vehicle Decision-Making and Control in Complex and Unconventional Scenarios—A Review

The development of autonomous vehicles (AVs) is becoming increasingly important as the need for reliable and safe transportation grows. However, in order to achieve level 5 autonomy, it is crucial that such AVs can navigate through complex and unconventional scenarios. It has been observed that currently deployed AVs, like human drivers, struggle the most in cases of adverse weather conditions, unsignalized intersections, crosswalks, roundabouts, and near-accident scenarios. This review paper provides a comprehensive overview of the various navigation methodologies used in handling these situations. The paper discusses both traditional planning methods such as graph-based approaches and emerging solutions including machine-learning based approaches and other advanced decision-making and control techniques. The benefits and drawbacks of previous studies in this area are discussed in detail and it is identified that the biggest shortcomings and challenges are benchmarking, ensuring interpretability, incorporating safety as well as road user interactions, and unrealistic simplifications such as the availability of accurate and perfect perception information. Some suggestions to tackle these challenges are also presented.

MITIGATING UNCERTAINTY IN CONCEPTUAL DESIGN USING OPERATIONAL SCENARIO SIMULATIONS: A DATA-DRIVEN EXTENSION OF THE EVOKE APPROACH

AbstractThe paper presents an approach where the iterative replication of Discrete Event Simulations on future operational scenarios is used to derive data-driven design merit functions. The presented contribution proposes an extension of the EVOKE (Early Value Oriented Design Exploration with Knowledge Maturity) approach determining when and how the experience-based judgment about maximization, minimization, optimization, and avoidance functions, correlating value drivers and quantified objectives, can be substituted by data-driven mathematical functions obtained by scenarios simulations. The approach is described through a simplified case concerning the development of autonomous electric vehicles to complement the public transport system in the city of Karlskrona in Sweden. The consideration of value drivers and quantified objectives presented is meant to support a preliminary screening of potential design configurations to support the definition of high-level product and system-related functional requirements, to be run before a more detailed conceptual design analysis.

Overweight and overload implementation in Apollo systems

Apollo is a high-performance, flexible architecture developed by Baidu, Kinglong, and a consortium of more than 40 companies for the purpose of accelerating the development, testing, and deployment of Autonomous Vehicles. This article first describes Apollo in terms of the functionality of the system and the interaction between its components, and then analyzes the conceptual architecture through system evolution, control and data flow, and concurrency, while illustrating the impact of the division of developer responsibilities on this. Then will go through a detailed overview of the concrete architecture of Baidu Apollo with working on mapping the source code from the Apollo Github Website to our architecture using the diagram drawing tool Scitools Understand. And divided into five parts, the report concludes with five major parts: the process from mapping code to diagram; concrete architecture explanation; unexpected dependencies discovery; subsystem analysis, and sequence diagrams presented in a more accurate and specific way. Finally, this report will propose a specific feature or enhancement to the current concrete architecture. The impact that this feature or enhancement can have on the non-functional requirements and stakeholders of the system will be explored based on the existing architecture and component interactions. In this regard, this report discusses two specific approaches that can be implemented and compares the two approaches through SAAM analysis to determine the best approach.

Predictive Maintenance of Lead-Acid Batteries with Sparse Vehicle Operational Data

Predictive maintenance aims to predict failures in components of a system, a heavy-duty vehicle in this work, and do maintenance before any actual fault occurs. Predictive maintenance is increasingly important in the automotive industry due to the development of new services and autonomous vehicles with no driver who can notice first signs of a component problem. The lead-acid battery in a heavy vehicle is mostly used during engine starts, but also for heating and cooling the cockpit, and is an important part of the electrical system that is essential for reliable operation. This paper develops and evaluates two machine-learning based methods for battery prognostics, one based on Long Short-Term Memory (LSTM) neural networks and one on Random Survival Forest (RSF). The objective is to estimate time of battery failure based on sparse and non-equidistant vehicle operational data, obtained from workshop visits or over-the-air readouts. The dataset has three characteristics: 1) no sensor measurements are directly related to battery health, 2) the number of data readouts vary from one vehicle to another, and 3) readouts are collected at different time periods. Missing data is common and is addressed by comparing different imputation techniques. RSF- and LSTM-based models are proposed and evaluated for the case of sparse multiple-readouts. How to measure model performance is discussed and how the amount of vehicle information influences performance.

Online Adaptive Generation of Critical Boundary Scenarios for Evaluation of Autonomous Vehicles

Testing and evaluation are critical steps in the development and deployment of autonomous vehicles (AVs). This paper aims to provide an online adaptive generation framework for critical boundary driving scenarios (CBDS) with flexible complexity and diversity. Both static scenarios and dynamic scenarios are comprehensively investigated based on real traffic flow data, where the generated static scenarios contain variables of drivable area, weather visibility, and road friction coefficient on autonomous vehicles, and the generated dynamic scenarios consider the effects of mixed traffic streams of AVs, human-driven vehicles, bicycles, and pedestrians. Two complexity models are proposed to characterize the complexity of static scenarios and dynamic scenarios separately, which help to construct feedback for the adaptive generation of CBDS. The scenario-based test is carried out on a joint simulation platform, and the multi-dimensional evaluation system, including safety, driving comfort, driving performance, and traffic coordination, is developed to assess the performance of AVs. Based on the proposed complexity models and multi-dimensional evaluation system, the generation of CBDS is transformed into online environmental parameter optimization. Moreover, the naturalistic driving scenarios generation, scenario complexity calculation, intelligent driving algorithm execution, and intelligent driving evaluation are all concatenated together to form a closed loop so as to adaptively generate critical boundary driving scenarios. Extensive simulations are conducted at the intersection with two different types of intelligent driving algorithms, which show the effectiveness of the proposed framework.

Real-Time Remote Control and Autonomous Navigation of a Two-Wheeled Robot with ROS and Raspberry Pi 4

Abstract: The development of autonomous vehicles has been the subject of considerable research and development in recent years. The project aims to develop a 2-wheeled autonomous car using a Robotic Operating System (ROS) with lidar and camera sensors as inputs. The car's control system will be designed to make decisions based on the information provided by these sensors, allowing it to navigate and avoid obstacles in real-time. The project will begin by selecting appropriate hardware components such as motors, batteries and sensors and integrating them into a functional prototype. Lidar and camera sensors will be used to detect and localize the position of the car, detect obstacles and determine their distance from the vehicle. The control system will be developed using ROS, an open source software framework for robot development, and will include algorithms for navigation, obstacle avoidance and decision making. The car will be programmed to follow a predetermined route. Finally, the autonomous car's performance will be evaluated through various tests, including its ability to follow a specific path and avoid obstacles. The results of these tests will be analyzed, and possible improvements will be identified. The outcome of this project is expected to provide a proof-of-concept for the development of autonomous vehicles using ROS with lidar and camera sensors as input. This technology has the potential to revolutionize the transportation industry by providing safer and more efficient transportation solutions

DATASET FOR POSITIONING AND TRACKING CARS AND PEDESTRIANS FROM UAV IMAGERY AND STATIC LIDAR

Abstract. The development of autonomous vehicles, both terrestrial and aerial ones (unmanned aerial system (UAS)), is causing the need of properly formulating appropriate solutions for ensuring a safe interaction between them, human beings and the infrastructures and environment in their operating area. To such aim, the knowledge of the positions of different platforms moving in the considered area is fundamental. GNSS (Global Navigation Satellite System) is by far the most used positioning technique in order to determine positions all over the world. Nevertheless, there are several conditions in which its use is unfortunately impossible or unreliable. Hence, different techniques, based on the use of sensors either mounted on the moving platforms or on an ad-hoc infrastructure, shall be used in order to determine the absolute and relative positions of the involved platforms. To this aim, this work proposes the use of vision, in particular from UAS imagery, static LiDAR (Light Detection and Ranging) and UWB (Ultra Wide-Band) transceivers, with initial encouraging results.

Safety Challenges and Analysis of Autonomous Electric Vehicle Development: Insights from On-Road Testing and Accident Reports

Autonomous electric vehicles (AEVs) hold great promise for the future of automotive engineering, but safety remains a significant challenge in their development and commercialization. Therefore, conducting a comprehensive analysis of AEV development and reported accidents is crucial. This paper reviews the levels of automation in AEVs, their disengagement frequencies, and on-road accident reports. According to the report, numerous manufacturers thoroughly tested AEVs across a distance of more than 3.9 million miles between 2014 and 2022. Disengagement frequencies vary among manufacturers, and approximately 65% of accidents during this period occurred while AEVs were operating in autonomous mode. Notably, the majority of accidents (90%) were caused by other road users, with only a small fraction (approximately 8%) directly attributed to AEVs. Enhancing AEVs' ability to detect and mitigate safety risks from external sources has the potential to significantly improve their safety. This paper provides valuable insights into AEV safety by emphasizing the importance of comprehensively understanding AEV development and reported accidents. Through the analysis of disengagement and accident reports, the study highlights the prevalence of passive accidents caused by other road users. Future research should concentrate on enabling AEVs to effectively detect and respond to safety risks originating from external sources to enhance AEV safety. Overall, this analysis contributes to the ongoing efforts in AEV development and provides guidance for strategies aimed at improving their safety features.

Vehicle Detection Using YOLOv5

Object detection is a computer vision method that allows for the identification and localization of specific classes of objects in images or videos. It goes beyond simple object classification and helps provide a better understanding of the object in question. Object detection has numerous applications, including automating business processes like inventory management in retail. It can detect objects that occupy between 2% and 60% of an image's area and clusters of objects as a single entity. Additionally, it can localize objects at high speeds, typically greater than 15 frames per second. Vehicle detection is a crucial component in the development of autonomous vehicles, enabling them to identify and perceive objects in their environment. It involves identifying and locating vehicles in image or video frames and has various applications in surveillance and security systems. There are different techniques and models for object detection, including traditional image processing methods and modern deep learning networks. Traditional methods like Viola- Jones, SIFT, and histogram of oriented gradients do not require historical data for training and are unsupervised. Popular image processing tools like OpenCV can be used for these techniques. On the other hand, modern deep learning networks like CNN, RCNN, YOLO, ResNet, RetinaNet, and MANet are supervised and efficient for object detection.

Advanced Scenario Generation for Calibration and Verification of Autonomous Vehicles

As driving scenarios and autonomous vehicles (AVs) become increasingly intricating, there is an increasing need for innovative frameworks that can enhance and test AV capabilities across diverse scenarios. At present, the design and validation for AVs predominantly rely on simulation software or real-road testing methodologies. However, these approaches possess inherent limitations, leading to inaccuracy, reduced efficiency, and potential hazards during the development process. This letter reports our first DHW (decentralized and hybrid workshop) on Scenarios Engineering (SE), that aims to calibrate and validate AV modules through advanced scenarios. Specifically, the DHW discusses combining SE with existing simulation software and real-road testing strategies, enabling developers to maximize the advantages of each approach while ensuring the safety and efficacy in the development of autonomous vehicles. Our findings demonstrate the benefits of enhancing realism of simulation scenarios in the aspects of both content and appearance. Furthermore, the use of scenarios engineering is explored in order to enhance the diversity and safety of real-road testing by means of integrating virtual and physical components.

Modelling communication-enabled traffic interactions.

A major challenge for autonomous vehicles is handling interactions with human-driven vehicles-for example, in highway merging. A better understanding and computational modelling of human interactive behaviour could help address this challenge. However, existing modelling approaches predominantly neglect communication between drivers and assume that one modelled driver in the interaction responds to the other, but does not actively influence their behaviour. Here, we argue that addressing these two limitations is crucial for the accurate modelling of interactions. We propose a new computational framework addressing these limitations. Similar to game-theoretic approaches, we model a joint interactive system rather than an isolated driver who only responds to their environment. Contrary to game theory, our framework explicitly incorporates communication between two drivers and bounded rationality in each driver's behaviours. We demonstrate our model's potential in a simplified merging scenario of two vehicles, illustrating that it generates plausible interactive behaviour (e.g. aggressive and conservative merging). Furthermore, human-like gap-keeping behaviour emerged in a car-following scenario directly from risk perception without the explicit implementation of time or distance gaps in the model's decision-making. These results suggest that our framework is a promising approach to interaction modelling that can support the development of interaction-aware autonomous vehicles.

Investigating pedestrian crossing decision with autonomous cars in virtual reality

With the development of autonomous vehicle (AV) technology, understanding how pedestrians interact with AVs is of increasing importance. In most field studies on pedestrian crossing behavior when encountering AVs, pedestrians were not permitted to physically cross the street due to safety restrictions. Instead, the physical crossing experience was replaced with indirect methods (e.g., by signalizing with gestures). We hypothesized that this lack of a physical crossing experience could influence the participants’ crossing behavior. To test this hypothesis, we adapted a reference study and constructed a crossing facility using a virtual reality (VR) simulation. In a controlled experiment, the participants encountered iterations of oncoming AVs. For each interaction, they were asked to either cross the street or signify their crossing decisions by taking steps at the edge of the street without crossing.Our study reveals that the lack of a physical crossing can lead to a significantly lower measured critical gap and perceived stress levels, thus indicating the need for detailed analysis when indirect methods are applied for future field studies.Practical Relevance: Due to safety requirements, experiments will continue to measure participants’ crossing behavior without permitting them to physically walk in front of an oncoming vehicle. Our study was the first attempt to reveal how this lack of crossing could potentially affect pedestrians’ behavior, and we obtained empirical evidence in support of our hypothesis, thus providing insights for future studies.

Adaptive Scheduling MAC Protocol in Underwater Acoustic Broadcast Communications for AUV Formation

With the rapid development of autonomous underwater vehicles (AUVs) and the continuous improvement of marine exploration requirements, AUV formation has emerged as a promising technique for performing underwater tasks with greater flexibility, adaptability, and scalability. To achieve AUV formation control, underwater acoustic communication networks (UACNs) are widely used to support information exchange among AUVs. Considering broadcast communications in fully connected mobile UACNs, existing medium access control (MAC) protocols face severe packet collision risk or low reuse efficiency. To improve the timeliness and reliability of the information exchange among AUVs, we develop an efficient scheduling-based adaptive broadcasting MAC (AB-MAC) protocol. The AB-MAC protocol is incompletely centralized, which can be reflected in two aspects. First, to ensure the consistency of the transmission schedule in the network, the AB-MAC protocol selects a control node for each frame to schedule the transmission of all nodes according to the updated network topology. The slot lengths and transmission sequences are changeable in each frame to avoid packet collisions and minimize frame lengths. Second, to further improve the efficiency and robustness of the protocol, all nodes adaptively adjust their transmission time by integrating the information contained in their previously received control packets and ordinary packets. The transmission scheduling problem is formulated as a combinatorial optimization problem, which can be solved by our proposed improved Genetic Algorithm. Numerical results demonstrate that the AB-MAC protocol not only performs well in terms of network throughput and average update interval but also has a high level of robustness even in low-quality underwater acoustic channels.

Computer Vision Application Analysis based on Object Detection

Object detection is a computer vision method that allows for the identification and localization of specific classes of objects in images or videos. It goes beyond simple object classification and helps provide a better understanding of the object in question. Object detection has numerous applications, including automating business processes like inventory management in retail [1]. It can detect objects that occupy between 2% and 60% of an image's area and clusters of objects as a single entity. Additionally, it can localize objects at high speeds, typically greater than 15 frames per second. Vehicle detection is a crucial component in the development of autonomous vehicles, enabling them to identify and perceive objects in their environment. It involves identifying and locating vehicles in image or video frames and has various applications in surveillance and security systems. There are different techniques and models for object detection, including traditional image processing methods and modern deep learning networks. Traditional methods like Viola-Jones, SIFT, and histogram of oriented gradients do not require historical data for training and are unsupervised. Popular image processing tools like OpenCV can be used for these techniques. On the other hand, modern deep learning networks like CNN, RCNN, YOLO, ResNet, RetinaNet, and MANet are supervised and efficient for object detection. Key Words: Deep learning, OpenCV, object detection

Learning naturalistic driving environment with statistical realism

For simulation to be an effective tool for the development and testing of autonomous vehicles, the simulator must be able to produce realistic safety-critical scenarios with distribution-level accuracy. However, due to the high dimensionality of real-world driving environments and the rarity of long-tail safety-critical events, how to achieve statistical realism in simulation is a long-standing problem. In this paper, we develop NeuralNDE, a deep learning-based framework to learn multi-agent interaction behavior from vehicle trajectory data, and propose a conflict critic model and a safety mapping network to refine the generation process of safety-critical events, following real-world occurring frequencies and patterns. The results show that NeuralNDE can achieve both accurate safety-critical driving statistics (e.g., crash rate/type/severity and near-miss statistics, etc.) and normal driving statistics (e.g., vehicle speed/distance/yielding behavior distributions, etc.), as demonstrated in the simulation of urban driving environments. To the best of our knowledge, this is the first time that a simulation model can reproduce the real-world driving environment with statistical realism, particularly for safety-critical situations.

Trajectory-based traffic observation of cooperation at a road narrowing

Abstract Understanding human interactions in today’s transportation system is a prerequisite for developing well-accepted cooperatively interacting autonomous vehicles. This paper is devoted to the two-sided narrow passage scenario and uses trajectory data to investigate drivers’ interaction behavior when encountering each other from opposite directions. Trajectory data of 209 encounters at a road narrowing were analyzed in terms of drivers’ approaching behavior and arrival order. The exploratory analysis has shown that in this specific location an informal traffic rule has developed: It was not the order of arrival but the direction of travel that primarily determined who passed the road narrowing first. This result shows that informal rules can influence drivers’ interaction behavior and should accordingly be considered in the development of autonomous vehicles to ensure safe and efficient encounters with human road users.

A precautionary approach to autonomous vehicles.

In this article, we defend an approach to autonomous vehicle ethics and policy based on the precautionary principle. We argue that a precautionary approach is warranted, given the significant scientific and moral uncertainties related to autonomous vehicles, especially higher-level ones. While higher-level autonomous vehicles may offer many important benefits to society, they also pose significant risks, which are not fully understood at this juncture. Risk management strategies traditionally used by government officials to make decisions about new technologies cannot be applied to higher-level autonomous vehicles because these strategies require accurate and reliable probability estimates concerning the outcomes of different policy options and extensive agreement about values, which are not currently available for autonomous vehicles. Although we describe our approach as precautionary, that does not mean that we are opposed to autonomous vehicle development and deployment, because autonomous vehicles offer benefits that should be pursued. The optimal approach to managing the risks of autonomous vehicles is to take reasonable precautions; that is, to adopt policies that attempt to deal with serious risks in a responsible way without depriving society of important benefits.

State of ADAS, Automation, and Connectivity

The vehicles are currently in a new stage of development, including advanced driver assistance, connectivity and automation. The timing and path of implementation of this technology remains uncertain due to continued technology development and changing consumer preferences. Although this new technology is still evolving, understanding the current automotive landscape helps to understand what the future holds. The Center for Automotive Research is dedicated to keeping the public informed about the current state of automotive technology. In 2017, 2019 and 2020, CAR released roadmaps outlining the current and likely future state of automotive technology. To ensure that the information contained in CAR's forecasts remains current and reflects the latest automotive technologies, trends and expectations, CAR conducts ongoing research and hosts industry roundtables. Industry roundtable experts include automakers, component vendors, technology vendors, data vendors, autonomous platform vendors and technology investors. Following the roundtable, the CAR developed an updated roadmap and white paper reflecting industry trends and expectations. This article summarizes key issues in the development and deployment of advanced driver assistance systems (ADAS), autonomous vehicles (AV), and vehicle connectivity technologies.

Dense reinforcement learning for safety validation of autonomous vehicles.

One critical bottleneck that impedes the development and deployment of autonomous vehicles is the prohibitively high economic and time costs required to validate their safety in a naturalistic driving environment, owing to the rarity of safety-critical events1. Here we report the development of an intelligent testing environment, where artificial-intelligence-based background agents are trained to validate the safety performances of autonomous vehicles in an accelerated mode, without loss of unbiasedness. From naturalistic driving data, the background agents learn what adversarial manoeuvre to executethrough a dense deep-reinforcement-learning (D2RL) approach, in which Markov decision processes are edited by removing non-safety-critical states and reconnecting critical ones so that the information in the training data is densified. D2RL enables neural networks to learn from densified information with safety-critical events and achieves tasks that are intractable for traditional deep-reinforcement-learning approaches. We demonstrate the effectiveness of our approach by testing a highly automated vehicle in both highway and urban test tracks with an augmented-reality environment, combining simulated background vehicles with physical road infrastructure anda real autonomous test vehicle. Our results show that the D2RL-trained agents can accelerate the evaluation process by multiple orders of magnitude (103 to 105 times faster). In addition, D2RL will enable accelerated testing and training with other safety-critical autonomous systems.