Cooperative Autonomous Driving Research Articles

Research on the Application of Vehicle-Road Cooperative Autonomous Driving

Research on the Application of Vehicle-Road Cooperative Autonomous Driving

Enhancing Transportation Safety with Infrastructure Cooperative Autonomous Driving System

Enhancing Transportation Safety with Infrastructure Cooperative Autonomous Driving System

Matching 5G Connected Vehicles to Sensed Vehicles for Safe Cooperative Autonomous Driving

Matching 5G Connected Vehicles to Sensed Vehicles for Safe Cooperative Autonomous Driving

Cooperative Autonomous Driving in Simulation

Autonomous driving is an area that has been growing in recent years. However, cars are unprepared to cooperate with others nearby, wasting resources and computational power. Thus, cooperative autonomous driving reveals its importance in the future. In this work-in-progress paper, we define, implement and test an architecture for a simulation environment where cooperative autonomous driving protocols can be tested. Additionally, a Manoeuvre Negotiation Protocol is implemented. This protocol will make an existing autonomous driving (AD) stack more resilient in real driving scenarios, improving its robustness and safety.

On Scalable Network Communication for Infrastructure-Vehicle Collaborative Autonomous Driving

Network communication is crucial for Infrastructure-Vehicle Cooperative Autonomous Driving (IVCAD), which can help to improve both driving safety and efficiency. Existing IVCAD systems either use RoadSide Units (RSUs) deployed sparsely in critical locations to extend the on-board sensing, or use RSUs deployed densely in a limited area to provide both sensing and decision-making for vehicles. For both scenarios, the numbers of RSUs and vehicles are small, and the network communication solutions are based on the existing wireline and cellular network infrastructures and their extensions. However, it remains unknown whether these existing network communication solutions are scalable when the IVCAD system is deployed city-widely with hundreds of thousands of RSUs and millions of vehicles. In this paper, we analyze the network communication requirements for different types of IVCAD systems, and show that when the numbers of RSUs and vehicles in the system are large, using the existing solutions to support IVCAD can incur very high extra network communication facility cost. We further discuss how to use wireless mesh networks formed by RSUs themselves to solve the network communication problem without any deployment of extra network communication facilities. To enable the RSU network, we provide a network communication framework based on network coding and highlight some research directions.

State Estimation and Motion Prediction of Vehicles and Vulnerable Road Users for Cooperative Autonomous Driving: A Survey

The recent progress in autonomous vehicle research and development has led to increasingly widespread testing of fully autonomous vehicles on public roads, where complex traffic scenarios arise. Along with these vehicles, partially autonomous vehicles, manually-driven vehicles, pedestrians, cyclists, and some animals can be present on the road, to which autonomous vehicles must react. This study focuses on a comprehensive survey of the literature on motion prediction and state estimation of vehicles and VRUs, which are essential for path planning and navigation functionalities of an autonomous vehicle. Motion prediction and state estimation methods utilize the vehicle’s own sensory perception capabilities and information obtained through cooperative perception from V2V and V2X connections. This survey summarizes the significant progress that has been made in both categories, discusses the most promising results to date and outlines critical research challenges that need to be overcome to achieve full autonomy, from an ego vehicle’s perspective in mixed traffic environments.

Cooperative Autonomous Driving for Urban Intersections Assisted by Vehicular Sensor Networks

The current autonomous driving vehicles cannot collaborate with each other due to the disconnection of sensors between vehicles, which leads to poor performance in traffic efficiency and safety. Recently, with the development of wireless intelligent sensors, vehicular sensor network (VSN) is emerging to provide information sharing between vehicles, so that cooperative autonomous driving (CAD) can be enabled. This paper especially investigates the VSN-assisted straight-going CAD at urban intersections under the rule of traffic signals. First, the function of VSN is presented, and the information sharing mechanism is designed. Then distributed stop-and-go strategies with maintaining the optimal safety spacing between adjacent vehicles are proposed for multiple traffic cases. With the proposed strategies, every vehicle can get an exact time to decelerate or accelerate at signalized intersections. Finally, we use the model predictive control (MPC) scheme to actuate our strategies, and it is demonstrated that vehicles perform well in both passage safety and efficiency as expected.

Communication Challenges in Infrastructure-Vehicle Cooperative Autonomous Driving: A Field Deployment Perspective

We have commercially deployed an infrastructure-vehicle cooperative autonomous driving system and verified its superiority over on-vehicle only autonomous driving systems. However, we have also identified network communication as the main technical roadblock for reliable cooperative autonomous driving. Hence, the goal of this article is to help the communications research community to better understand the real-world communication challenges in cooperative autonomous driving, as well as to introduce practical solutions for establishing a good baseline for further research and development efforts. Specifically, we introduce the real-world network communication challenges: first, the current mobile network bandwidth is constraining the uploading of raw sensing data, which is crucial for cloud-based applications such as deep learning model training and high-definition map generation and so on. Second, the network latency jitters remain high for a considerable portion of a vehicle's trip, greatly impacting the reliability and safety of the operations of autonomous vehicles. To address these two challenges, we have developed, deployed, and verified two practical solutions. First, a sensing compression strategy to cope with the network bandwidth challenge. Second, an adaptive fusion engine to cope with the latency variation challenge.

Optimizing Information Freshness for MEC-Enabled Cooperative Autonomous Driving

Fully automated vehicles deployed with high computational/perceptive capabilities will soon become a reality. Such capabilities enable the cooperation among vehicles and the realization of interacting autonomous driving systems. Edge computing has emerged to provide a plethora of computational services to reduce network latency. Applications at the edge that apply analytics on the sensory data are therefore indispensable for self-driving vehicles. We consider in this paper a network that interconnects vehicles to an edge server at a roadside unit. Each vehicle extracts multiple information by sampling multiple processes and sends them to the corresponding edge application. To make timely decisions, “fresh” information needs to be offloaded, processed, and delivered back to vehicles; in this context, we adopt a new metric called Age of Information (AoI) that has been lately used to measure the freshness of information. We seek to jointly schedule vehicles’ transmission of information and schedule information processing at the edge to minimize the AoI of all processes. We mathematically formulate the problem and prove its NP-Hardness. To overcome this hardness, we propose a logic-based Benders decomposition to divide the problem into a master and several subproblems. Then, we present an exact polynomial-time solution for the subproblems, a scalable heuristic for the master, and devise a valid yet efficient Benders cut. We implement the system simulation on the well-known traffic simulator SUMO and compare the decomposition with CPLEX branch-and-cut; Although the problem is highly intricate, our method finds a near-optimal solution (maximum deviation is 7% from optimal solution) with a speedup that reaches 95%. We study the system performance by varying different system parameters.

The Business Case for Infrastructure-Vehicle Cooperative Autonomous Driving

This article is the first to discuss the business case for infrastructure-vehicle cooperative autonomous driving (IVCAD), or the utilization of infrastructure-side sensing and computing capabilities to empower autonomous vehicles to be smarter and to make the whole transportation system more efficient. IVCAD is a rising technology area under immense debate, with most of the criticism around the economic rationality of infrastructure investments. In this article, through our real-world deployment experiences, we establish the business case for IVCAD and discuss why IVCAD is more economical and efficient compared to the on-vehicle only autonomous driving approach.

Robustness and Adaptability of Reinforcement Learning-Based Cooperative Autonomous Driving in Mixed-Autonomy Traffic

Building autonomous vehicles (AVs) is a complex problem, but enabling them to operate in the real world where they will be surrounded by human-driven vehicles (HVs) is extremely challenging. Prior works have shown the possibilities of creating inter-agent cooperation between a group of AVs that follow a social utility. Such altruistic AVs can form alliances and affect the behavior of HVs to achieve socially desirable outcomes. We identify two major challenges in the co-existence of AVs and HVs. First, social preferences and individual traits of a given human driver, e.g., selflessness and aggressiveness are unknown to an AV, and it is almost impossible to infer them in real-time during a short AV-HV interaction. Second, contrary to AVs that are expected to follow a policy, HVs do not necessarily follow a stationary policy and therefore are extremely hard to predict. To alleviate the above-mentioned challenges, we formulate the mixed-autonomy problem as a multi-agent reinforcement learning (MARL) problem and propose a decentralized framework and reward function for training cooperative AVs. Our approach enables AVs to learn the decision-making of HVs implicitly from experience, optimizes for a social utility while prioritizing safety and allowing adaptability; robustifying altruistic AVs to different human behaviors and constraining them to a safe action space. Finally, we investigate the robustness, safety and sensitivity of AVs to various HVs behavioral traits and present the settings in which the AVs can learn cooperative policies that are adaptable to different situations.

도심자율협력주행 시스템의 V2V 통신에 대한 기능안전 평가시나리오 개발 방법론에 관한 연구

도심자율협력주행 시스템의 V2V 통신에 대한 기능안전 평가시나리오 개발 방법론에 관한 연구

대열주행을 위한 V2V 통신 기반 자율협력주행 알고리즘 개발

대열주행을 위한 V2V 통신 기반 자율협력주행 알고리즘 개발

5G NR-V2X: Toward Connected and Cooperative Autonomous Driving

5G New Radio (NR) is touted as a pivotal enabling technology for the genuine realization of connected and cooperative autonomous driving. Despite numerous research efforts in recent years, a systematic overview on the role of 5G NR in future connected autonomous communication networks is missing. To fill this gap and to spark more future research, this article introduces the technology components of 5G NR and discusses the evolution from existing cellular vehicle-to-everything (V2X) technology toward NR-V2X. We primarily focus on the key features and functionalities of the physical layer, sidelink communication and its resource allocation, architecture flexibility, security and privacy mechanisms, and precise positioning techniques. Moreover, we envisage and highlight the potential of machine learning for further performance enhancement in NR-V2X services. Lastly, we show how 5G NR can be configured to support advanced V2X use cases.

A Comparison of Trajectory Planning and Control Frameworks for Cooperative Autonomous Driving

Abstract This work considers the cooperative trajectory-planning problem along a double lane change scenario for autonomous driving. In this paper, we develop two frameworks to solve this problem based on distributed model predictive control (MPC). The first approach solves a single nonlinear MPC problem. The general idea is to introduce a collision cost function in the optimization problem at the planning task to achieve a smooth and bounded collision function, and thus to prevent the need to implement tight hard constraints. The second method uses a hierarchical scheme with two main units: a trajectory-planning layer based on mixed-integer quadratic program (MIQP) computes an on-line collision-free trajectory using simplified motion dynamics, and a tracking controller unit to follow the trajectory from the higher level using the nonlinear vehicle model. Connected and automated vehicles (CAVs) sharing their planned trajectories lay the foundation of the cooperative behavior. In the tests and evaluation of the proposed methodologies, matlab-carsim cosimulation is utilized. carsim provides the high-fidelity model for the multibody vehicle dynamics. matlab-carsim conjoint simulation experiments compare both approaches for a cooperative double lane change maneuver of two vehicles moving along a one-way three-lane road with obstacles.

Multiclock Constraint System Modelling and Verification for Ensuring Cooperative Autonomous Driving Safety

CADS (cooperative autonomous driving systems) are software-intensive and safety-critical reactive systems and give great promise to our daily life, but system errors may not be identified in the design stage until the implement stage, and the cost to correct them will be more expensive later than the early stage. For designing trustworthy autonomous software systems, we have to deal with multiclock constraint models. SysML (System Modeling Language) meets increasing adoption in order to carry out system-level modelling and verification against abstract representations, but it suffers from semantic ambiguities in the design of safety-critical autonomous systems. The main objective is to investigate methods for coping with the design and analysis models simultaneously and to achieve semantic consistency based on mathematical foundations and formal model transformation. In this paper, we propose a method to combine the requirement modelling process with analysis process together for CADS safety and reliability guarantee. Firstly, we extend SysML metamodels and construct SysML profile for the CADS domain that could improve modelling correctness and enhance reusability. An instantiated CADS model has been designed by means of adopting a profile containing different key functional and nonfunctional attributes and behaviors. Secondly, we define formal syntax and semantic notations for modelling elements in the SysML state machine diagram and show transformation rules between the state machine diagram and the CCSL (Clock Constraint Specification Language) model. Semantic preservation is also proved using the bisimulation relation between them for rigorous mapping correctness. Thirdly, a cooperative autonomous overtaking driving case study on the highway scenario is used for illustration, and we use the tool TimeSquare to simulate CCSL specification execution traces at the system design stage.

Latest Technology Trends and Issues for Realizing Cooperative Autonomous Driving

Latest Technology Trends and Issues for Realizing Cooperative Autonomous Driving

Cooperative Autonomous Driving Oriented MEC-Aided 5G-V2X: Prototype System Design, Field Tests and AI-Based Optimization Tools

Vehicle-to-Everything (V2X) requirements from cooperative autonomous driving can be characterized as ultra-reliable, low latency, high traffic, and high mobility. These requirements introduce great challenges in the air interface and protocol stack design, resource allocation, network deployment, and all the way up to mobile (or multi-access) edge computing (MEC), cloud and application layer. In this paper, we present a cooperative autonomous driving oriented MEC-aided 5G-V2X prototype system design and the rationale behind the design choices. The prototype system is developed based on a next-generation radio access network (NG-RAN) experimental platform, a cooperative driving vehicle platoon, and an MEC server providing high definition (HD) 3D dynamic map service. Field tests are conducted and the results demonstrate that the combination of 5G-V2X, MEC and cooperative autonomous driving can be pretty powerful. Considering the remaining challenges in the commercial deployment of 5G-V2X networks and future researches, we propose two artificial intelligence (AI) based optimization tools. The first is a deep-learning-based tool called deep spatio-temporal residual networks with a permutation operator (PST-ResNet). By providing city-wide user and network traffic prediction, PST-ResNet can help to reduce the capital expense (CAPEX) and operating expense (OPEX) costs of commercial 5G-V2X networks. The second is a swarm intelligence based optimization tool called subpopulation collaboration based dynamic self-adaption cuckoo Search (SC-DSCS), which can be widely used to solve complex optimization problems in future researches. The effectiveness of proposed optimization tools is verified by real-world data and benchmark functions.

Development and Field Application of an Amphibious Scrubbing/Suction Dredging Machine with Cylindrical Rotating Brush and Turbidity Barrier

An amphibious scrubbing/suction dredging machine with cylindrical rotating brush, housing, and turbidity barrier was newly-developed to remove both sediments with about 10 cm thickness and periphyton attached on various structures in urban water-circulating systems through the scrubbing, suction, and dredging processes. Based on the field application and long-term monitoring, the increase in both suspended solids (SS) and turbidity of water during the scrubbing, suction, and dredging processes was negligible (p>0.05). In some cases, the turbidity of water initially increased, however, the turbidity was stabilized within 20 minutes from the start of dredging processes. The concentration changes in TN and TP of water were not statistically different (p>0.05) before and after the scrubbing, suction, and dredging processes, indicating that benthic nutrients released from sediments were not significantly diffused, and were not supposed to cause significant water pollution. Also, water treatment facilities along with an amphibious scrubbing/suction dredging machine could be more effective since the removal of contaminant loadings through the scrubbing, suction, and dredging processes was much greater than that through simple coagulation/precipitation processes. Finally, GPS-based realtime tracking and operation program have been developed and applied in various urban water-circulating systems, and development of driver cooperative autonomous driving system is in progress to eliminate the need for manual driving of an amphibious scrubbing/suction dredging machine. Key words: Urban Water-Circulating System, Suspended Solids, Turbidity, Sediments, Nutrients, Cylindrical Rotating Brush, Driver Cooperative Autonomous Driving System, Amphibious Scrubbing/Dredging Machine

Cooperative Autonomous Driving: A Mirror Neuron Inspired Intention Awareness and Cooperative Perception Approach

In this article, we present the concept of cooperative autonomous driving using mirror neuron inspired intention awareness and cooperative perception, whose primary benefit is to enable cooperative autonomous driving in a more general approach rather than complicated rule-based approaches. The cooperative perception can provide information on the upcoming traffic situations ahead, even beyond line-of-sight and field-of-view. From a control perspective, a spatial map for navigation planning is extended up to the boundary of connected vehicles in a see-through manner. By leveraging this augmented perception capability, a better driving decision can be accomplished in terms of traffic flow efficiency and safety improvement. For this purpose, we propose a mirror neuron inspired intention awareness algorithm along with planning and control methods for cooperative autonomous driving. We demonstrate the feasibility of our proposals through simulations and experiments on the road with a cooperative lane changing scenario.