Showing papers on "Collision avoidance published in 2021"

Risk assessment based collision avoidance decision-making for autonomous vehicles in multi-scenarios

Abstract: In this paper, we proposed a new risk assessment based decision-making algorithm to guarantee collision avoidance in multi-scenarios for autonomous vehicles. A probabilistic-model based situation assessment module using conditional random field was proposed to assess the risk level of surrounding traffic participants. Based on the assessed risk from the situation assessment module, a collision avoidance strategy with driving style preferences (e.g., aggressive or conservative) was proposed to meet the demands of different drivers or passengers. Finally, we conducted experiments in Carla (car learning to act) to evaluate our developed collision avoidance decision-making algorithm in different scenarios. The results show that our developed method was sufficiently reliable for autonomous vehicles to avoid collisions in multi-scenarios with different driving style preferences. Our developed method with adjustable driving style preferences to meet the demand of different consumers would improve drivers’ acceptance of autonomous vehicles.

Collision Avoidance in Pedestrian-Rich Environments With Deep Reinforcement Learning

Abstract: Collision avoidance algorithms are essential for safe and efficient robot operation among pedestrians. This work proposes using deep reinforcement (RL) learning as a framework to model the complex interactions and cooperation with nearby, decision-making agents, such as pedestrians and other robots. Existing RL-based works assume homogeneity of agent properties, use specific motion models over short timescales, or lack a principled method to handle a large, possibly varying number of agents. Therefore, this work develops an algorithm that learns collision avoidance among a variety of heterogeneous, non-communicating, dynamic agents without assuming they follow any particular behavior rules. It extends our previous work by introducing a strategy using Long Short-Term Memory (LSTM) that enables the algorithm to use observations of an arbitrary number of other agents, instead of a small, fixed number of neighbors. The proposed algorithm is shown to outperform a classical collision avoidance algorithm, another deep RL-based algorithm, and scales with the number of agents better (fewer collisions, shorter time to goal) than our previously published learning-based approach. Analysis of the LSTM provides insights into how observations of nearby agents affect the hidden state and quantifies the performance impact of various agent ordering heuristics. The learned policy generalizes to several applications beyond the training scenarios: formation control (arrangement into letters), demonstrations on a fleet of four multirotors and on a fully autonomous robotic vehicle capable of traveling at human walking speed among pedestrians.

Path planning and collision avoidance for autonomous surface vehicles I: a review

Abstract: Autonomous surface vehicles are gaining increasing attention worldwide due to the potential benefits of improving safety and efficiency. This has raised the interest in developing methods for path planning that can reduce the risk of collisions, groundings, and stranding accidents at sea, as well as costs and time expenditure. In this paper, we review guidance, and more specifically, path planning algorithms of autonomous surface vehicles and their classification. In particular, we highlight vessel autonomy, regulatory framework, guidance, navigation and control components, advances in the industry, and previous reviews in the field. In addition, we analyse the terminology used in the literature and attempt to clarify ambiguities in commonly used terms related to path planning. Finally, we summarise and discuss our findings and highlight the potential need for new regulations for autonomous surface vehicles.

Self-Configuring Robot Path Planning With Obstacle Avoidance via Deep Reinforcement Learning

Abstract: This letter proposes a hybrid control methodology to achieve full body collision avoidance in anthropomorphic robot manipulators. The proposal improves classical motion planning algorithms by introducing a Deep Reinforcement Learning (DRL) approach trained ad hoc for performing obstacle avoidance, while achieving a reaching task in the operative space. More specifically, a switching mechanism is enabled whenever a condition of proximity to the obstacles is met, thus conferring to the dual-mode architecture a self-configuring capability in order to cope with objects unexpectedly invading the workspace. The proposal has been finally tested relying on a realistic robot manipulator simulated in a V-REP environment.

Automatic ship collision avoidance using deep reinforcement learning with LSTM in continuous action spaces

Abstract: This paper presents an automatic collision avoidance algorithm for ships using a deep reinforcement learning (DRL) in continuous action spaces. Obstacle zone by target (OZT) is used to compute an area where a collision will happen in the future based on dynamic information of ships. Agents of DRL detects the approach of multiple ships using a virtual sensor called the grid sensor. Agents learned collision avoidance maneuvering through Imazu problem, which is a scenario set of ship encounter situations. In this study, we propose a new approach for collision avoidance with a longer safe passing distance using DRL. We develop a novel method named inside OZT that expands OZT to improve the consistency of learning. We redesign the network using the long short-term memory (LSTM) cell and carried out training in continuous action spaces to train a model with longer safe distance than the previous study. The bow cross range in collision detection proposed in this paper is effective to COLREGs-compliant collision avoidance. The trained model has passed all scenarios of Imazu problem. The model is also validated by a test scenario which includes more ships than each scenario of Imazu problem.

Improved YOLOv4 Marine Target Detection Combined with CBAM

Abstract: Marine target detection technology plays an important role in sea surface monitoring, sea area management, ship collision avoidance, and other fields. Traditional marine target detection algorithms cannot meet the requirements of accuracy and speed. This article uses the advantages of deep learning in big data feature learning to propose the YOLOv4 marine target detection method fused with a convolutional attention module. Marine target detection datasets were collected and produced and marine targets were divided into ten categories, including speedboat, warship, passenger ship, cargo ship, sailboat, tugboat, and kayak. Aiming at the problem of insufficient detection accuracy of YOLOv4’s self-built marine target dataset, a convolutional attention module is added to the YOLOv4 network to increase the weight of useful features while suppressing the weight of invalid features to improve detection accuracy. The experimental results show that the improved YOLOv4 has higher detection accuracy than the original YOLOv4, and has better detection results for small targets, multiple targets, and overlapping targets. The detection speed meets the real-time requirements, verifying the effectiveness of the improved algorithm.

Formation Control With Collision Avoidance Through Deep Reinforcement Learning Using Model-Guided Demonstration

Abstract: Generating collision-free, time-efficient paths in an uncertain dynamic environment poses huge challenges for the formation control with collision avoidance (FCCA) problem in a leader–follower structure. In particular, the followers have to take both formation maintenance and collision avoidance into account simultaneously. Unfortunately, most of the existing works are simple combinations of methods dealing with the two problems separately. In this article, a new method based on deep reinforcement learning (RL) is proposed to solve the problem of FCCA. Especially, the learning-based policy is extended to the field of formation control, which involves a two-stage training framework: an imitation learning (IL) and later an RL. In the IL stage, a model-guided method consisting of a consensus theory-based formation controller and an optimal reciprocal collision avoidance strategy is designed to speed up training and increase efficiency. In the RL stage, a compound reward function is presented to guide the training. In addition, we design a formation-oriented network structure to perceive the environment. Long short-term memory is adopted to enable the network structure to perceive the information of obstacles of an uncertain number, and a transfer training approach is adopted to improve the generalization of the network in different scenarios. Numerous representative simulations are conducted, and our method is further deployed to an experimental platform based on a multiomnidirectional-wheeled car system. The effectiveness and practicability of our proposed method are validated through both the simulation and experiment results.

Path planning and collision avoidance for autonomous surface vehicles II: a comparative study of algorithms

Abstract: Artificial intelligence is an enabling technology for autonomous surface vehicles, with methods such as evolutionary algorithms, artificial potential fields, fast marching methods, and many others becoming increasingly popular for solving problems such as path planning and collision avoidance. However, there currently is no unified way to evaluate the performance of different algorithms, for example with regard to safety or risk. This paper is a step in that direction and offers a comparative study of current state-of-the art path planning and collision avoidance algorithms for autonomous surface vehicles. Across 45 selected papers, we compare important performance properties of the proposed algorithms related to the vessel and the environment it is operating in. We also analyse how safety is incorporated, and what components constitute the objective function in these algorithms. Finally, we focus on comparing advantages and limitations of the 45 analysed papers. A key finding is the need for a unified platform for evaluating and comparing the performance of algorithms under a large set of possible real-world scenarios.

A Detailed Survey and Future Directions of Unmanned Aerial Vehicles (UAVs) with Potential Applications

Abstract: Recently, unmanned aerial vehicles (UAVs), also known as drones, have gained widespread interest in civilian and military applications, which has led to the development of novel UAVs that can perform various operations. UAVs are aircraft that can fly without the need of a human pilot onboard, meaning they can fly either autonomously or be remotely piloted. They can be equipped with multiple sensors, including cameras, inertial measurement units (IMUs), LiDAR, and GPS, to collect and transmit data in real time. Due to the demand for UAVs in various applications such as precision agriculture, search and rescue, wireless communications, and surveillance, several types of UAVs have been invented with different specifications for their size, weight, range and endurance, engine type, and configuration. Because of this variety, the design process and analysis are based on the type of UAV, with the availability of several control techniques that could be used to improve the flight of the UAV in order to avoid obstacles and potential collisions, as well as find the shortest path to save the battery life with the support of optimization techniques. However, UAVs face several challenges in order to fly smoothly, including collision avoidance, battery life, and intruders. This review paper presents UAVs’ classification, control applications, and future directions in industry and research interest. For the design process, fabrication, and analysis, various control approaches are discussed in detail. Furthermore, the challenges for UAVs, including battery charging, collision avoidance, and security, are also presented and discussed.

Intelligent UAV Swarm Cooperation for Multiple Targets Tracking

Abstract: With the advantages of easy deployment and flexible usage, Unmanned Aerial Vehicle (UAV) has advanced the Multi-Target Tracking (MTT) applications. The UAV-MTT system has great potentials to execute dull, dangerous, and critical missions for frontier defense and security. A key challenge in UAV-MTT is how to coordinate multiple UAVs to track diverse invading targets accurately and consecutively. In this paper, we propose a UAV swarm-based cooperative tracking architecture to systematically improve the UAV tracking performance. We design an intelligent UAV swarm-based cooperative algorithm for consecutive target tracking and physical collision avoidance. Moreover, we design an efficient cooperative algorithm to predict the trajectory of invading targets accurately. Our simulation results demonstrate that the swarm behaviors stay stable in realistic scenarios with perturbing obstacles. Compared with state-of-the-art solutions such as the matched deep Q-network, our algorithms can increase tracking accuracy by 60%, reduce tracking delay by 23%, and achieve physical collision-avoidance during the tracking process.

Where to go Next: Learning a Subgoal Recommendation Policy for Navigation in Dynamic Environments

Abstract: Robotic navigation in environments shared with other robots or humans remains challenging because the intentions of the surrounding agents are not directly observable and the environment conditions are continuously changing. Local trajectory optimization methods, such as model predictive control (MPC), can deal with those changes but require global guidance, which is not trivial to obtain in crowded scenarios. This letter proposes to learn, via deep Reinforcement Learning (RL), an interaction-aware policy that provides long-term guidance to the local planner. In particular, in simulations with cooperative and non-cooperative agents, we train a deep network to recommend a subgoal for the MPC planner. The recommended subgoal is expected to help the robot in making progress towards its goal and accounts for the expected interaction with other agents. Based on the recommended subgoal, the MPC planner then optimizes the inputs for the robot satisfying its kinodynamic and collision avoidance constraints. Our approach is shown to substantially improve the navigation performance in terms of number of collisions as compared to prior MPC frameworks, and in terms of both travel time and number of collisions compared to deep RL methods in cooperative, competitive and mixed multiagent scenarios.

Multiple-Target Surrounding and Collision Avoidance With Second-Order Nonlinear Multiagent Systems

Abstract: This article proposes an equal-distance surrounding control method for second-order nonlinear multiagent systems (MASs) to encircle multiple moving targets with guaranteed collision avoidance. First, a distributed estimator is developed to approximate the center of moving targets. Then, an adaptive distributed control law is designed for the MASs to accomplish equal-distance surrounding collaboratively. In particular, conditions for assuring asymptotical stability for the closed-loop MAS are derived. Finally, experimental results with unmanned surface vessels are reported to substantiate the effectiveness of the proposed coordinated surrounding control method.

Finite-Time Bearing-Only Formation Tracking of Heterogeneous Mobile Robots With Collision Avoidance

Abstract: This brief proposes a bearing-only collision-free formation coordination strategy for networked heterogeneous robots, where each robot only measures the relative bearings of its neighbors to achieve cooperation Different from many existing studies that can only guarantee global asymptotic stability (ie, the formation can only be formed over an infinite settling period), a gradient-descent control protocol is designed to make the robots achieve a target formation within a given finite time The stability of the multi-robot system is guaranteed via Lyapunov theory, and the convergence time can be defined by users Moreover, we also present sufficient conditions for collision avoidance Finally, a simulation case study is provided to verify the effectiveness of the proposed approach

A Survey of Driving Safety With Sensing, Vehicular Communications, and Artificial Intelligence-Based Collision Avoidance

Abstract: Accurately discovering hazards and issuing appropriate warnings to drivers in advance or performing autonomous control is the core of the Collision Avoidance (CA) system used to solve traffic safety problems. More comprehensive environmental awareness, diversified communication technologies, and autonomous control can make the CA system more accurate and effective, thereby improving driving safety. In addition, the assistance of Artificial Intelligence (AI) technology can make the CA system adapt to the environment and facilitate fast and accurate decisions. Considering the current lack of a thorough survey of driving safety with sensing, vehicular communications, and AI-based collision avoidance, in this paper, we survey existing researches for state-of-the-art data-driven CA techniques. Firstly, we discuss the major steps of CA and key research issues. For each step, we review the existing enabling techniques and research methods for CA in detail, including sensing and vehicular communication for safe driving, as well as CA algorithm design. Particularly, we present a comparison between the most common AI algorithms for different functions in the CA system. Testbeds and projects for CA are summarized next. Finally, several open challenges and future research directions are also outlined.

Adversarial Reinforcement Learning Framework for Benchmarking Collision Avoidance Mechanisms in Autonomous Vehicles

Abstract: With the rapidly growing interest in autonomous navigation, the body of research on motion planning and collision avoidance techniques has enjoyed an accelerating rate of novel proposals and developments. However, the complexity of new techniques and their safety requirements render the bulk of current benchmarking frameworks inadequate, thus leaving the need for efficient comparison techniques unanswered. This work proposes a novel framework based on deep reinforcement learning for benchmarking the behavior of collision avoidance mechanisms under the worst-case scenario of dealing with an optimal adversarial agent, trained to drive the system into unsafe states. We describe the architecture and flow of this framework as a benchmarking solution, and demonstrate its efficacy via a practical case study of comparing the reliability of two collision avoidance mechanisms in response to adversarial attempts to cause collisions.

Integrated Steering and Differential Braking for Emergency Collision Avoidance in Autonomous Vehicles

Abstract: Controlling the lateral dynamics of an autonomous vehicle confronting a sudden obstacle requires optimal use of tires’ force capacities. In these situations, autonomous steering may not be able to respond fast enough to prevent collision or instability. This paper presents an integrated controller for autonomous vehicles, capable of suitably reacting to emergency situations when a sudden obstacle appears on the road. The proposed controller employs differential braking conservatively when needed, to produce an additional yaw moment, thereby improving a vehicle’s lateral agility and responsiveness without endangering vehicle stability. A longitudinal controller is also designed to track a desired longitudinal velocity. Model predictive control (MPC) method is used for developing a combined path planning and tracking controller with a hierarchical structure that prioritizes (1) collision avoidance, (2) vehicle stability, and (3) path tracking. The effectiveness of the proposed integrated MPC controller is evaluated by simulating an experimentally validated CarSim model to demonstrate the controller’s capability in preventing instability and collisions.

Collision Risk Inference System for Maritime Autonomous Surface Ships Using COLREGs Rules Compliant Collision Avoidance

Abstract: Maritime autonomous surface ships (MASS) need to be sufficiently safe to gain commercial acceptance. Collision avoidance strategies in such MASS should comply with the International Regulations for Preventing Collision at the Sea (COLREGs). According to the COLREGs, collision risk assessment, which determines the optimal positioning and timing via all available means appropriate to the prevailing circumstances and conditions, is crucial for preventing collisions. However, existing collision risk assessment methods do not consider all vital factors for the COLREGs rules compliant collision avoidance. We propose a collision risk inference system for MASS that complies with COLREGs vital rules for collision avoidance as follows: 1) actions to avoid collision are defined according to the degree of danger, and a suitable response distance is determined; 2) a collision risk index according to the enlarged ship domain based on the designated response distance by each level is set; 3) all vital factors of the COLREGs rules compliant collision avoidance are extracted as the data when the ship domain enlarged by each level is overlapped; 4) the collision risk inference system is developed by learning extracted data via the adaptive neuro fuzzy inference system. In contrast to existing research, the proposed system considers all vital variables in the COLREGs rules compliant collision avoidance guidelines, thereby improving the timings and positionings of the potential collision warning. Consequently, it could secure more time for decision making to take necessary collision prevention action.