Search

Autonomous Vehicles (AV) are expected to bring considerable benefits to society, such as traffic optimization and accidents reduction. They rely heavily on advances in many Artificial Intelligence (AI) approaches and techniques. However, while some researchers in this field believe AI is the core element to enhance safety, others believe AI imposes new challenges to assure the safety of these new AI-based systems and applications. In this non-convergent context, this paper presents a systematic literature review to paint a clear picture of the state of the art of the literature in AI on AV safety. Based on an initial sample of 4870 retrieved papers, 59 studies were selected as the result of the selection criteria detailed in the paper. The shortlisted studies were then mapped into six categories to answer the proposed research questions. An AV system model was proposed and applied to orient the discussions about the SLR findings. As a main result, we have reinforced our preliminary observation about the necessity of considering a serious safety agenda for the future studies on AI-based AV systems. [ABSTRACT FROM AUTHOR]

In the recent past, platooning evolved into an attractive cooperative driving technology, broadly discussed in research and practice. Vehicles in platoons use cooperative adaptive cruise control to drive at close distances to each other. Platooning (i) increases the capacity of the street by a factor of 2; (ii) reduces the fuel consumption and emissions by up to 20%; and (iii) has social implications as it increases driver comfort and safety. As platooning research progresses, platooning coordination becomes a major research focus. The coordination of platoons, including the assignment of vehicles to platoons, the management of inter- and intra-platoon interactions, and the coordination of interactions with other vehicles is an important step towards an effective usage of platooning in practice. Based on a literature review of 1,600 papers, this survey provides an overview of state of the art in platooning coordination research for both cars and trucks. In this paper, we present a novel taxonomy for platooning coordination and classify existing approaches. We use the results of the literature review to discuss challenges and outline avenues for future work such as multi-objectiveness and individualisation. [ABSTRACT FROM AUTHOR]

The technical maturity of autonomous driving enables the discussion of beneficial use cases to leverage its full potential. In this paper, we target one such use case: Platooning is the efficient convoying of vehicles by making use of self-driving capabilities and inter-vehicle communication. Many advantages arise from grouping vehicles in platoons with a small inter-vehicle distance, such as energy savings, congestion reduction, and safety improvements. However, due to the diversity of involved stakeholders, numerous objectives have to be balanced to leverage the full potential of platooning. Furthermore, these objectives also depend on various factors that influence their optimization. The vast majority of existent literature only targets a subset of related objectives and underlying factors. This paper provides an overview which categorizes objectives and influencing factors. Additionally, metrics for the evaluation of objective attainment are proposed. [ABSTRACT FROM AUTHOR]

In this paper, a new framework for path tracking is proposed through learning to drive like human beings. Firstly, the imitation algorithm (behavior cloning) is adopted to initialize the deep reinforcement learning (DRL) algorithm through learning the professional drivers’ experience. Secondly, a continuous, deterministic, model free deep reinforcement learning algorithm is adopted to optimize our DRL model on line through trial and error. By combining behavior cloning and deep reinforcement learning algorithms, the DRL model can learn an effective policy quickly for path tracking using some easy-to-measure vehicle state parameters and environment information as inputs. Actor-Critic structure is adopted in the DRL algorithm. In order to speed up the convergence rate of the DRL model and improve the learning effect, we propose a dual actor networks structure for the two different action outputs (steering wheel angle and vehicle speed), and a chief critic network is built to guide the updating process of dual actor networks at the same time. Based on this dual actor networks structure, we can pick out some more important state information as state inputs for different action outputs. Besides, a kind of reward mechanism is also presented for autonomous driving. Finally, simulation training and experiment test are carried out, and the results confirm that the framework proposed in this paper is more than data efficient than the original algorithm, and the trained DRL model can track the reference path with accuracy and has the generalization ability for different roads. [ABSTRACT FROM AUTHOR]

Recently, new research activities have emerged for controlling traffic. Since future vehicles will travel autonomously and communicate with their surrounding environment, they will then be able to negotiate the right-of-way at intersections [cooperative intersection management (CIM)] as well as reserve their itinerary (road reservation). Both concepts, i.e., CIM and road reservation, are very promising for relieving traffic congestion. The scope of this paper is to prevent deadlock under real-time conditions, by taking advantage of these two concepts. This paper presents an appropriate graph to model the network of intersections and a sufficient condition for obtaining deadlock-free traffic with a specified route from their origin to their destination. In addition, it proposes a hierarchical approach in which a network server, intersection servers, and vehicles contribute to improving the traffic condition. Simulations were performed on a network of 25 interconnected intersections, as well as, on a real urban network to prove the effectiveness of the model. The results are presented and discussed. [ABSTRACT FROM AUTHOR]

End-to-end steering estimation is one of the important deep regression tasks. However, driving datasets are always imbalanced on the distribution of the steering value, which makes end-to-end learning models perform bad steering estimation accuracy, especially for the sharp steering value estimation, which finally, leads to an imbalanced training problem. In this paper, the essential reason for the imbalanced training problem for the deep regression task is first revealed as the gradient disharmony. To hedge the gradient disharmony, a general gradient balanced mechanism for the deep regression task is proposed in this paper. Meanwhile, the general gradient balanced mechanism is applied for the steering estimation task. According to the steering distribution, a novel Adaptive Gradient Balanced Mechanism (AGBM) is proposed to hedge the gradient disharmony based on the Laplace distribution with the general gradient balanced mechanism theory. Further, a new loss function embedding with AGBM is proposed to balance the gradient disharmony for the training process. AGBM makes the process of gradient hedging with adaptive style without parameters finetuning. Based on five driving datasets, four experiments are conducted for demonstrations. The experiment results show the AGBM enables the end-to-end models to perform the best accuracy of the steering estimation, including the sharp steering samples. Meanwhile, AGBM is generalized with different end-to-end learning models. [ABSTRACT FROM AUTHOR]

In this paper, we propose a new state representation method, called encoding sum and concatenation (ESC), to describe the environment observation for decision-making in autonomous driving. Unlike existing state representation methods, ESC is applicable to the situation where the number of surrounding vehicles is variable and eliminates the need for manually pre-designed sorting rules, leading to higher representation ability and generality. The proposed ESC method introduces a feature neural network (NN) to encode the real-valued feature of each surrounding vehicle into an encoding vector, and then adds these vectors up to obtain the representation vector of the set of surrounding vehicles. Then, a fixed-dimensional and permutation-invariance state representation can be obtained by concatenating the set representation with other variables, such as indicators of the ego vehicle and road. By introducing the sum-of-power mapping, this paper has further proved that the injectivity of the ESC state representation can be guaranteed if the output dimension of the feature NN is greater than the number of variables of all surrounding vehicles. This means that the ESC representation can be used to describe the environment and taken as the inputs of learning-based policy functions. Experiments demonstrate that compared with the fixed-permutation representation method, the policy learning accuracy based on ESC representation is improved by 62.2%. [ABSTRACT FROM AUTHOR]

Recent advances in smart connected vehicles and Intelligent Transportation Systems (ITS) are based upon the capture and processing of large amounts of sensor data. Modern vehicles contain many internal sensors to monitor a wide range of mechanical and electrical systems and the move to semi-autonomous vehicles adds outward looking sensors such as cameras, lidar, and radar. ITS is starting to connect existing sensors such as road cameras, traffic density sensors, traffic speed sensors, emergency vehicle, and public transport transponders. This disparate range of data is then processed to produce a fused situation awareness of the road network and used to provide real-time management, with much of the decision making automated. Road networks have quiet periods followed by peak traffic periods and cloud computing can provide a good solution for dealing with peaks by providing offloading of processing and scaling-up as required, but in some situations latency to traditional cloud data centres is too high or bandwidth is too constrained. Cloud computing at the edge of the network, close to the vehicle and ITS sensor, can provide a solution for latency and bandwidth constraints but the high mobility of vehicles and heterogeneity of infrastructure still needs to be addressed. This paper surveys the literature for cloud computing use with ITS and connected vehicles and provides taxonomies for that plus their use cases. We finish by identifying where further research is needed in order to enable vehicles and ITS to use edge cloud computing in a fully managed and automated way. We surveyed 496 papers covering a seven-year timespan with the first paper appearing in 2013 and ending at the conclusion of 2019. [ABSTRACT FROM AUTHOR]

This paper investigates the cyber-security problem for autonomous vehicles under sensor attacks. In particular, a model-based framework is proposed which can detect sensor attacks and identify their sources in order to achieve the secure localization of self-driving vehicles. To ensure robustness of the vehicle against cyber-attacks, sensor redundancy is introduced, that is to deploy multiple sensors, each of which provides real-time pose observations of the vehicle. A bank of attack detectors is developed to capture anomalies in each sensor measurement, which is a combination of an extended Kalman filter (EKF) and a cumulative sum (CUSUM) discriminator. EKFs are employed to estimate the vehicle position and orientation recursively, while each CUSUM discriminator is designed to analyze the residual generated by its combined EKF to detect the possible deviation of the sensor measurement from the expected pose derived according to the mathematical model of the vehicle. To monitor the inconsistency amongst multiple sensor measurements, an auxiliary detector is introduced which fuses observations from multiple sensors. Based on the results of all the detectors, a rule-based isolation scheme is developed to identify the source anomalous sensor. The effectiveness of our proposed framework has been demonstrated on real vehicle data. [ABSTRACT FROM AUTHOR]

Identifying and evaluating the potential risks in the surrounding environment is critical for intelligent vehicles’ safety and user experience. This paper provides a comprehensive overview of the state-of-the-art research on the surrounding vehicles’ lane change maneuver prediction and detection. First, various driver behavior modeling and classification methods are reviewed and analyzed, which gives a general understanding of what the lane change maneuver is and how to predict or detect the lane change maneuver. Next, the primary sensing devices equipped on intelligent vehicles and their impacts on lane change inference systems are discussed. Then, a series of representative research works in recent years are selected, introduced, and compared regarding their input feature selection, inference algorithms, and performance evaluation methods. Finally, some potential future research directions are proposed. This paper aims to help the relevant researchers and institutions summarize the current studies on the surrounding vehicles’ lane change maneuver inference and recognize its future development directions. [ABSTRACT FROM AUTHOR]

Traditional vehicle localization methods use information from a GPS, inertial navigation, and an odometer, but a GPS is often limited by availability in urban areas for its sensitivity to terrain and interference. Aiming at the above problems, this paper presents a method of intersection localization using a traffic lights’ map. In particular, traffic lights with significant visual characteristics in the urban environment are used as landmarks. The localization accuracy of the autonomous vehicle at intersections is improved by combining the location information of traffic lights provided by a high-precision map. The whole scheme of the localization system is designed and the sensors used are determined. First, the coordinates of sensors are established, respectively, and the transformation and rotation are calibrated. Then, the state model and measurement model of the vehicle vision system are established. Combined with the location and height information of the traffic lights provided by the high-precision map, the extended Kalman filter is used to fuse the vision detection results of the traffic lights with the inertial measurement unit information. The experiments demonstrate that the method proposed in this paper improves the lateral localization accuracy and the accuracy of the vehicle’s yaw angle. [ABSTRACT FROM AUTHOR]

Academic research in the field of autonomous vehicles has reached high popularity in recent years related to several topics as sensor technologies, V2X communications, safety, security, decision making, control, and even legal and standardization rules. Besides classic control design approaches, Artificial Intelligence and Machine Learning methods are present in almost all of these fields. Another part of research focuses on different layers of Motion Planning, such as strategic decisions, trajectory planning, and control. A wide range of techniques in Machine Learning itself have been developed, and this article describes one of these fields, Deep Reinforcement Learning (DRL). The paper provides insight into the hierarchical motion planning problem and describes the basics of DRL. The main elements of designing such a system are the modeling of the environment, the modeling abstractions, the description of the state and the perception models, the appropriate rewarding, and the realization of the underlying neural network. The paper describes vehicle models, simulation possibilities and computational requirements. Strategic decisions on different layers and the observation models, e.g., continuous and discrete state representations, grid-based, and camera-based solutions are presented. The paper surveys the state-of-art solutions systematized by the different tasks and levels of autonomous driving, such as car-following, lane-keeping, trajectory following, merging, or driving in dense traffic. Finally, open questions and future challenges are discussed. [ABSTRACT FROM AUTHOR]

Autonomous driving of multi-lane vehicle platoons have attracted significant attention in recent years due to their potential to enhance the traffic-carrying capacity of the roads and produce better safety for drivers and passengers. This paper proposes a distributed motion planning algorithm to ensure safe overtaking of autonomous vehicles in a dynamic environment using the Artificial Potential Field method. Unlike the conventional overtaking techniques, autonomous driving strategies can be used to implement safe overtaking via formation control of unmanned vehicles in a complex vehicle platoon in the presence of human-operated vehicles. Firstly, we formulate the overtaking problem of a group of autonomous vehicles into a multi-target tracking problem, where the targets are dynamic. To model a multi-vehicle system consisting of both autonomous and human-operated vehicles, we introduce the notion of velocity difference potential field and acceleration difference potential field. We then analyze the stability of the multi-lane vehicle platoon and propose an optimization-based algorithm for solving the overtaking problem by placing a dynamic target in the traditional artificial potential field. A simulation case study has been performed to verify the feasibility and effectiveness of the proposed distributed motion control strategy for safe overtaking in a multi-lane vehicle platoon. [ABSTRACT FROM AUTHOR]

Connected autonomous vehicle technology is conductive to promoting the transition from traditional merging control (e.g., ramp metering) to automated merging control. This paper proposes a platoon-based hierarchical merging control algorithm for on-ramp vehicles to achieve automated merging control under connected traffic environment. The proposed algorithm optimizes merging maneuvers of on-ramp vehicles to smooth their merging trajectories without frequent decelerations or stops at the end of the ramp, and to minimize disruption to the mainline traffic in the merging zone. A tactical layer controller is designed to select pre-target merging gaps for on-ramp vehicles, in which the future motion (i.e., acceleration and deceleration) of mainline vehicles is considered through the grey prediction model. An operational layer controller is constructed based on model predictive control to adjust the speed of on-ramp vehicles in advance, and controls on-ramp vehicles to merge into the pre-target merging gaps under state constraints (i.e., safe headway, maximum speed and so on). Through numerical simulation, the effectiveness of the proposed algorithm is validated under different merging scenarios. It is shown that on-ramp vehicles smoothly merge into the mainline within the pre-target merging gap at the same speed as adjacent mainline vehicles. Compared with the baseline merging control algorithm, the proposed algorithm significantly reduces both fuel consumption and travel time of on-ramp vehicles, and improves passenger comfort. [ABSTRACT FROM AUTHOR]

Autonomous Driving (AD) introduces dramatic changes to the way we travel. This emerging technology has the potential to impact the transportation sector across a wide array of categories including safety, efficiency, congestion, legislation, and travel behavior. In this survey, we review the main issues involved in AD as discussed in the literature, and shed light on topics that we consider requiring further development based on Google Scholar and the Web of Science (WoS) data base. The paper also provides the results of research trends related to Autonomous Driving based on analysis of the number of search results listed in google trends. According to our research, the fields of Vehicle-to-Vehicle (V2V) and Vehicle-to-Cloud (V2C) networking are of higher interest due to the technological gaps and standardization processes. In addition, cyber and security research is in acceleration due to its importance. [ABSTRACT FROM AUTHOR]

Delay and lag deteriorate path tracking accuracy and system stability. If not properly compensated, they can cause instability or limit the driving speed of autonomous vehicles. This paper presents a preview steering control design considering both communication delay and steering lag to achieve accurate, smooth, and computationally efficient path tracking for highly automated vehicles. Two strategies are adopted for the delay: forward state predictor and delay augmentation. The steering lag is approximated by a first-order lag system. The path tracking problem with delay and lag is solved by the preview control theory. The resulted controller is in an analytical form and is computationally efficient for online implementation. We also analyze the system stability and closed-loop responses in both the time and frequency domain. The control is implemented on an automated vehicle platform and tested inside Mcity and on open roads. Experimental results showed lower tracking errors and significantly improved stability margin compared to the controls ignoring delay and lag. [ABSTRACT FROM AUTHOR]

Autonomous Vehicles (AVs) are a significant part of Vehicular Adhoc NETwork (VANET) as they increase transportation accessibility. However, the presence of unpredictably sized potholes on road surfaces hampers the comfort and safety of autonomous navigation. Existing pothole avoidance mechanisms cannot dynamically adapt in unpredictable environments and do not comfort the traveler well in VANET. This paper proposes a novel Smart Pothole-Avoidance Strategy (SPAS) for safe navigation in a pothole-intensive environment. Potholes are avoided using the Deep Deterministic Policy Gradient (DDPG) algorithm as it performs best in continuous action space tasks and has a faster convergence speed. A Hybrid Recognition Model using the Speech and Gesture mechanism (HRM-SG) is proposed in this paper to collect the traveler’s real-time audio and visual feedback for the DDPG reward function. The received feedback aids in fine-tuning the model to avoid the pothole efficiently than the previous pothole. Traveler’s feedback is coupled with the vehicle’s sensor data and used to decide the time, speed, and angle at which lane change and speed change are executed. Finally, the SPAS continuously optimizes lane change parameters in VANET to achieve maximal traveler comfort during the operation. The result analysis indicates that SPAS achieves a 10-15% improvement in the accuracy of pothole avoidance, 10-12% higher comfort, and 8-10% faster convergence than the existing state-of-the-art techniques. [ABSTRACT FROM AUTHOR]

Vehicle localization in highway tunnels is challenging for autonomous vehicle navigation. The drift error increases because GPS signals from satellites cannot be received inside a highway tunnel. Therefore, map-aided localization is essential for correcting the drift error. However, the environment around the tunnel is composed mainly of elliptical walls, and there are few unique feature points for map matching, unlike the case outdoors. In this paper, we propose a tunnel facility–based precise vehicle localization method for highway tunnels using 3D LIDAR. For vehicle localization, a point landmark map storing the center points of the tunnel facilities and a probability distribution map storing the probability distributions of the lane markings is used. In the proposed method, the tunnel facility points are extracted using 3D LIDAR, and position estimation is conducted using an extended Kalman filter-based navigation filter. The proposed localization algorithm was verified through experiments using actual highway driving data. The experimental results verify that tunnel facility–based vehicle localization yields precise results in various situations. [ABSTRACT FROM AUTHOR]

Pedestrian Detection is the most critical module of an Autonomous Driving system. Although a camera is commonly used for this purpose, its quality degrades severely in low-light night time driving scenarios. On the other hand, the quality of a thermal camera image remains unaffected in similar conditions. This paper proposes an end-to-end multimodal fusion model for pedestrian detection using RGB and thermal images. Its novel spatio-contextual deep network architecture is capable of exploiting the multimodal input efficiently. It consists of two distinct deformable ResNeXt-50 encoders for feature extraction from the two modalities. Fusion of these two encoded features takes place inside a multimodal feature embedding module (MuFEm) consisting of several groups of a pair of Graph Attention Network and a feature fusion unit. The output of the last feature fusion unit of MuFEm is subsequently passed to two CRFs for their spatial refinement. Further enhancement of the features is achieved by applying channel-wise attention and extraction of contextual information with the help of four RNNs traversing in four different directions. Finally, these feature maps are used by a single-stage decoder to generate the bounding box of each pedestrian and the score map. We have performed extensive experiments of the proposed framework on three publicly available multimodal pedestrian detection benchmark datasets, namely KAIST, CVC-14, and UTokyo. The results on each of them improved the respective state-of-the-art performance. A short video giving an overview of this work along with its qualitative results can be seen at https://youtu.be/FDJdSifuuCs. Our source code will be released upon publication of the paper. [ABSTRACT FROM AUTHOR]

Most existing studies on autonomous intersection management (AIM) often focus on algorithms to accommodate conflicts among vehicles by assuming that the entrance lane and the exit lane of vehicles are exogenous inputs. This paper shows that allowing entrance lanes and exit lanes to be optimized can significantly improve traffic efficiency. In particular, this paper proposes “all-direction” lanes, where left-turn, through, and right-turn traffic is all allowed at the same lane. We develop two methods for optimizing entering time (i.e., when to enter the intersection) and route choice decisions (i.e., entrance lane and exit lane), including the sliding-time-window-based global optimum (GO-STW) and the first-come-first-served method with optimal route choices (FCFS-R). The developed lane-based methods can be formulated as mixed integer linear programming (MILP) problems, which can be solved using the CPLEX solver. A heuristic is further adopted to solve the MILP model in a timely manner, which illustrates the potential real-time applicability of the proposed method. Numerical analysis is conducted to examine performance and effectiveness of the proposed methods and heuristic. We found that the optimization of lane/route choices is often more critical than entering time. [ABSTRACT FROM AUTHOR]

This paper presents a hierarchical Green-Light Approach Speed (h-GLAS) strategy for controlling heavy-duty trucks traveling through urban/suburban environments, where future intersection signal phase and timing (SPaT) information is non-deterministic. Through vehicle-to-infrastructure communication, past SPaT information is sent to the vehicle and is used to forecast future predictions of the signal timing using a Gaussian Process (GP) model. The h-GLAS strategy uses the predicted SPaT information to generate an efficient desired velocity profile that navigates through intersections where the probability of a green light is maximized. This velocity profile is tracked by a convex model predictive controller (MPC) that simultaneously minimizes mechanical energy expenditure and braking effort over its prediction horizon. Downstream from the MPC, we implement a command governor (CG) that adjusts the MPC output by the minimum amount necessary to maintain safe vehicle following and ensure that the vehicle stops at all red lights. Using a proprietary medium-fidelity Simulink model provided by Volvo Trucks, we characterize the h-GLAS strategy’s performance over a real suburban route, consisting of 10 actuated signalized intersections, using actual SPaT information provided by the NC Department of Traffic (DOT). Simulation results demonstrate a 30-43% reduction in fuel consumption, as compared to a baseline control strategy, which is attributable primarily to avoiding the massive energy losses associated with braking. Furthermore, we evaluate the computational efficiency of our approach by assessing average simulation execution times. [ABSTRACT FROM AUTHOR]

The automotive industry has witnessed an increasing level of development in the past decades; from manufacturing manually operated vehicles to manufacturing vehicles with a high level of automation. With the recent developments in Artificial Intelligence (AI), automotive companies now employ blackbox AI models to enable vehicles to perceive their environment and make driving decisions with little or no input from a human. With the hope to deploy autonomous vehicles (AV) on a commercial scale, the acceptance of AV by society becomes paramount and may largely depend on their degree of transparency, trustworthiness, and compliance with regulations. The assessment of the compliance of AVs to these acceptance requirements can be facilitated through the provision of explanations for AVs’ behaviour. Explainability is therefore seen as an important requirement for AVs. AVs should be able to explain what they have ‘seen’, done, and might do in environments in which they operate. In this paper, we provide a comprehensive survey of the existing work in explainable autonomous driving. First, we open by providing a motivation for explanations by highlighting the importance of transparency, accountability, and trust in AVs; and examining existing regulations and standards related to AVs. Second, we identify and categorise the different stakeholders involved in the development, use, and regulation of AVs and elicit their AV explanation requirements. Third, we provide a rigorous review of previous work on explanations for the different AV operations (i.e., perception, localisation, planning, vehicle control, and system management). Finally, we discuss pertinent challenges and provide recommendations including a conceptual framework for AV explainability. This survey aims to provide the fundamental knowledge required of researchers who are interested in explanation provisions in autonomous driving. [ABSTRACT FROM AUTHOR]

The complex traffic and road environment pose considerable challenges to the accuracy, timeliness, and adaptive ability of connected and autonomous vehicles (CAVs) in making driving decisions. This paper uses vehicle collaboration and integrates the adaptive learning capabilities of machine learning and the interpretation capabilities of expert systems (ESs) in a unified architecture to form a hybrid autonomous driving guidance system, which not only solves the “bottleneck” of knowledge acquisition during the construction of expert systems but also solves the “black box” phenomenon of machine learning in the decision-making process. First, an autonomous driving strategy based on deep reinforcement learning (DRL) is proposed for CAVs to make decisions and extract corresponding rules. Next, we design an ES knowledge base expansion method including rule extraction, rule sharing, and rule test. Particularly, vehicular blockchain is adopted to ensure user privacy and data security during the rule-sharing process. Third, hybrid autonomous driving guidance combining ES and machine learning is proposed for CAVs to make accurate and efficient decisions in different driving environments. Once the strategy is well trained, it can effectively guide CAVs to cope with the complex traffic environment. Extensive simulations validate the performance of our proposal in terms of decision-making accuracy, effectiveness, and safety. [ABSTRACT FROM AUTHOR]

This paper is about obtaining stable vehicle platooning by using Cooperative Adaptive Cruise Control when the communication is unreliable and suffers from message losses. We model communication losses as independent random events and we propose an original design for the cooperative controller, which effectively mitigates the effects of the losses. Our design explicitly takes into account the stochastic nature of the losses by considering both the average evolution of the system and the stochastic variations around it. The control design promotes both plant stability and string stability of the average error dynamics by an $\mathscr {H}_\infty $ approach, while minimizing the variance of the trajectories around their average. We show by simulations that the proposed controller is able to compensate losses even for high loss probabilities. [ABSTRACT FROM AUTHOR]

Autonomous vehicles must be comprehensively evaluated before deployed in cities and highways. However, most existing evaluation approaches for autonomous vehicles are static and lack adaptability, so they are usually inefficient in generating challenging scenarios for tested vehicles. In this paper, we propose an adaptive evaluation framework to efficiently evaluate autonomous vehicles in adversarial environments generated by deep reinforcement learning. Considering the multimodal nature of dangerous scenarios, we use ensemble models to represent different local optimums for diversity. We then utilize a nonparametric Bayesian method to cluster the adversarial policies. The proposed method is validated in a typical lane-change scenario that involves frequent interactions between the ego vehicle and the surrounding vehicles. Results show that the adversarial scenarios generated by our method significantly degrade the performance of the tested vehicles. We also illustrate different patterns of generated adversarial environments, which can be used to infer the weaknesses of the tested vehicles. [ABSTRACT FROM AUTHOR]

Safety and efficiency are important priorities in transportation engineering. The new research directions in intelligent transportation systems focus on autonomy and collaboration to improve safety and efficiency. In this paper, we consider the problem of conflict resolution at an unsignalized intersection. Collision detection and avoidance are addressed based on an optimal strategy that uses collaboration between the vehicles. Lagrangian function is used to obtain closed-form solutions. This approach allows analysis of the collaboration process in terms of the cost and the time to resolve conflicts. Several approaches are used for weight assignments including reciprocal and unilateral actions. In this paper, reciprocal collaboration is defined as collaboration with equal costs. It is proven that the minimum cost to resolve any conflict corresponds to equal weights but not to reciprocal collaboration. On the other hand, reciprocal collaboration results in the minimum time to resolve a conflict. The proposed methods present effective solutions to the problem. Results are validated through simulation. [ABSTRACT FROM AUTHOR]

This paper considers platoons with an arbitrary number of linear vehicles and derives conditions on the vehicle dynamics for ensuring time-headway spacing and for guaranteeing collision avoidance. It is shown that for reaching these control aims it is necessary and sufficient that the controlled vehicles are externally positive with respect to both the velocity and the distance output and have a quantified delay. The requirement on the platoon dynamics are transformed into four objectives for the vehicle controller design. Three control configurations are evaluated with respect to these objectives to find systematic design steps. Numerical examples illustrate the results. [ABSTRACT FROM AUTHOR]

The higher social attention and negotiation among road users while crossing the road is as much a challenge for self-driving cars as it is for pedestrians. Self-driving cars in their current state are not able to understand cues from other road users and are rather reactive to pedestrian behavior, which may result overall in a slower traffic flow. In this paper, a vehicle–pedestrian negotiation model is proposed describing the processing and exchange of negotiation cues from both parties in order to speed up the traffic flow. The motion strategy for the vehicle approaching the pedestrian is formulated in order to negotiate its best chance to pass first, a process that closely mimics the common scenarios of everyday negotiation on roads. Simulation results show an improvement in the overall travel time of the vehicles as compared with the current best practice behavior (always stop) of autonomous vehicles. The cost-benefit analysis of negotiation among both parties is also discussed in this paper. [ABSTRACT FROM AUTHOR]

Tile-based reservation intersection control for autonomous vehicles has the potential to reduce intersection delays beyond optimized traffic signals. A major question in implementing reservations is the underdetermined problem of resolving conflicting reservation requests. Previous work studied prioritizing requests by first come first served or holding auctions at intersections, but the possibilities are infinite. Furthermore, although selfish routing behavior could affect the benefits of the reservation prioritization, reservation control has not been studied with user equilibrium routing due to its microsimulation definition. This paper addresses these issues by presenting an integer program formulation of the conflict point simplification of reservations. The feasible region is transformed, resulting in a more tractable integer program on conflict regions for dynamic traffic assignment. Because the integer program is NP-hard we present a polynomial-time heuristic. Finally, we demonstrate the potential utility of this heuristic by demonstrating objective functions that reduce travel time and energy consumption on a city network. [ABSTRACT FROM PUBLISHER]

Autonomous vehicles (AV) have advanced considerably over the past decade and their potential to reduce road accidents is without equal. That said, the evolution towards fully automated driving will be accompanied by new and unfamiliar risks. The deployment of AVs hinges on the premise that they are considerably safer than human drivers. However, the ability of manufacturers, insurers and regulators to quantifiably demonstrate this risk reduction, relative to humans, presents a major barrier. Based on accident rates, it will likely take hundreds of millions of autonomous miles to derive statistically meaningful results. This paper addresses this issue and proposes a novel means of quantifying AV accident risks by benchmarking against a more familiar and quantifiable risk – Human Behaviour. This method is used to proactively quantify AV safety relative to human drivers. Currently, anomalous driving behaviour stems from human susceptibilities such as fatigue or aggression. We exploit this observation and explore AV driving behaviour where driving anomalies are symptoms of technology errors. The comparative behaviours of AV and safe human driving can be used to measure AV accident risk. An end-to-end model AV is simulated using Convolutional Neural Networks (CNN) to compare human and AV driving behaviours. Using a machine learning technique called Gaussian Processes (GP), contextual driving anomalies are detected, the frequency and severity of which are used to derive a risk score. This paper offers a starting point for addressing the challenges surrounding AV risk modelling. [ABSTRACT FROM AUTHOR]

Imbalanced training is a challenge in the field of autonomous driving. For the steering estimation task, imbalanced training is the core reason that an end-to-end model cannot estimate sharp steering value well. Inspired by researches on the steering estimation, this paper proposes a novel loss function to train a high-performance end-to-end model for handling the imbalanced training problem, which is named SteeringLoss. Firstly, the imbalanced distribution of the steering value for driving datasets is analyzed, which is similar to the double long-tailed distribution. Secondly, with the feature of distribution, this paper designs a cost-sensitive loss function step by step, the new loss function can improve the impact of the sharp steering value while maintaining the impact of the small steering value. Thirdly, three typical steering estimation models are established with SteeringLoss for demonstration, including the CNN model, the CNN-LSTM model and the 3DCNN-LSTM model. Finally, three experiments are designed for SteeringLoss: Experiment I demonstrates SteeringLoss can avoid imbalanced training problem; Experiment II discusses the finetuning principle of SteeringLoss and gives the basic guideline for using SteeringLoss; Experiment III shows the results on different typical end-to-end steering estimation models, which shows effectiveness of SteeringLoss for all the models and gives different solutions for the steering estimation. Moreover, the SteeringLoss is suitable for the imbalanced training with similar distributions of datasets besides end-to-end steering estimation, which indicates the potential value of the research on SteeringLoss in the future. [ABSTRACT FROM AUTHOR]

Autonomous vehicles need to solve the road keeping problem and the existing solutions based on reinforcement learning are mainly implemented in the simulators. The key of transferring the well-trained models to the real world is bridging the gaps between the simulator scenarios and the real scenarios. In this paper, we propose a method called ReinforcementDriving which explores navigation skills and trajectories from simulator for full-sized road keeping. Based on the real scenario, a driving simulator is firstly established to train an intelligent driving agent. The well-trained ReinforcementDriving agent is evaluated in a real-world scenario. We compare our work with human driving, optimal control-based tracking methods and other reinforcement learning-based lane following methods. The results demonstrate that the ReinforcementDriving system can effectively achieve lane keeping in a realistic scenario with satisfactory running time and lateral accuracy. [ABSTRACT FROM AUTHOR]

The rapid advancement of sensor technologies and artificial intelligence are creating new opportunities for traffic safety enhancement. Dashboard cameras (dashcams) have been widely deployed on both human driving vehicles and automated driving vehicles. A computational intelligence model that can accurately and promptly predict accidents from the dashcam video will enhance the preparedness for accident prevention. The spatial-temporal interaction of traffic agents is complex. Visual cues for predicting a future accident are embedded deeply in dashcam video data. Therefore, the early anticipation of traffic accidents remains a challenge. Inspired by the attention behavior of humans in visually perceiving accident risks, this paper proposes a Dynamic Spatial-Temporal Attention (DSTA) network for the early accident anticipation from dashcam videos. The DSTA-network learns to select discriminative temporal segments of a video sequence with a Dynamic Temporal Attention (DTA) module. It also learns to focus on the informative spatial regions of frames with a Dynamic Spatial Attention (DSA) module. A Gated Recurrent Unit (GRU) is trained jointly with the attention modules to predict the probability of a future accident. The evaluation of the DSTA-network on two benchmark datasets confirms that it has exceeded the state-of-the-art performance. A thorough ablation study that assesses the DSTA-network at the component level reveals how the network achieves such performance. Furthermore, this paper proposes a method to fuse the prediction scores from two complementary models and verifies its effectiveness in further boosting the performance of early accident anticipation. [ABSTRACT FROM AUTHOR]

Deep learning (DL) brings autonomous vehicles (AVs) close to reality. However, the witness of many safety issues has raised a big concern about the reliability of AVs. To solve this problem, much research has been done to test deep learning-driven AVs. Generally, once a test input is produced, a developer needs to manually check its expected output. However, there often exists massive unlabeled test data (e.g., raw context traces in the real world). It is impractical to manually label all test inputs. Despite some works on automatic generation of test oracles, they are either task-specific or constrained to synthetic inputs. In this paper, we present a general and extensible framework, DeepSuite, to mitigate the manual effort of generating test oracles. The intuition behind is that not all test inputs are equally worth labelling. With limited testing budget, it is desirable to label a test suite with high diversity and a reasonable size. Due to the large search space, to optimize such test suites is of great challenge. To address it, DeepSuite employs a three-phase optimization method (i.e., selection, crossover, and mutation) to iteratively select representative but non-redundant test suites. Such conflicting profit/cost objectives are attained through a genetic algorithm with a well-defined multi-objective fitness function. In the experiments, we first show that the diversity of tests can be revealed by test criteria. Then, experiments on three widely-used datasets demonstrated the effectiveness of DeepSuite in generating test suites with competitive testing coverage and 68.42% smaller size, which greatly improves the data collection efficiency of testing DL-driven autonomous vehicles. [ABSTRACT FROM AUTHOR]

The rapid development of 6G can help to bring autonomous driving closed to the reality. Drivers and passengers will have more time for work and leisure spending in the vehicles, further generating a lot of data requirements. However, edge resources from small base stations are insufficient to match the wide variety of services of the future vehicular networks. Besides, due to the high-speed nature of the vehicles, users have to switch the connections among different base stations, whereas such way will cause external latency during the data request. Therefore, it is vital to enable the local cache of vehicle users to realize the reliable autonomous driving. In this paper, we consider caching the contents in the local cache, small base station, and edge server. In practice, the request preference of some single users may be different from a whole region. To maximize the efficiency of content cache, we design a strategy that uses reinforcement learning algorithm to optimize cache schemes on different devices. The experimental results demonstrate that our strategy can enhance the cache hit ratio by 10%-20% compared with the well-known counterparts. [ABSTRACT FROM AUTHOR]

The growth of autonomous vehicles, ridesharing systems, and self-driving technology will bring a shift in the way ride hailing platforms plan out their services. However, these advances in technology coupled with road congestion, environmental concerns, fuel usage, vehicles emissions, and the high cost of the vehicle usage have brought more attention to better utilize the use of vehicles and their capacities. In this paper, we propose a novel distributed multi-hop ride-sharing (MHRS) algorithm that uses deep reinforcement learning to learn optimal vehicle dispatch and matching decisions by interacting with the external environment. By allowing customers to transfer between vehicles, i.e., ride with one vehicle for some time and then transfer to another one, MHRS helps in attaining 30% lower cost and 20% more efficient utilization of fleets, as compared to the ride-sharing algorithms. This flexibility of multi-hop feature gives a seamless experience to customers and ride-sharing companies, and thus improves ride-sharing services. [ABSTRACT FROM AUTHOR]

In the era of Connected and Autonomous Vehicles, platooning has the potential to increase roadway capacity and reduce energy consumption. However, as vehicles try to form platoons they may also expend extra energy. Also, depending on its position within a platoon, the energy savings of each vehicle can be different. Thus, optimizing and quantifying the savings that may be gained from platooning is challenging. In this paper, we develop a simulation-optimization framework to tackle this challenge of quantifying energy savings from platooning. Our optimization model determines vehicle-to-platoon assignments given the current location, speed, and destination of all the vehicles and platoons on the freeway. The simulation model takes these platooning decisions from the optimization model and implements them. Vissim is used to simulate the actions taken by all the vehicles and platoons and capture the energy expended by each vehicle over its entire trip duration. To quantify the energy savings, the system is simulated with and without platooning. That is, the optimization model is turned off when assessing the performance of the system without platooning. In addition to the simulation-optimization framework, an accurate energy consumption model is developed in this paper, which is inspired by the work of Tadakuma and colleagues. The energy consumption model utilizes a hybrid prediction formula for aerodynamic drag reduction in multi-vehicle formations unifying both physical mechanisms and existing empirical study data. Our results show that a system wide savings of about 3% can be realized over 160 kilometers when platoons are formed strategically. [ABSTRACT FROM AUTHOR]

Reinforcement learning has gradually demonstrated its decision-making ability in autonomous driving. Reinforcement learning is learning how to map states to actions by interacting with environment so as to maximize the long-term reward. Within limited interactions, the learner will get a suitable driving policy according to the designed reward function. However there will be a lot of unsafe behaviors during training in traditional reinforcement learning. This paper proposes a RL-based method combined with RL agent and Monte Carlo tree search algorithm to reduce unsafe behaviors. The proposed safe reinforcement learning framework mainly consists of two modules: risk state estimation module and safe policy search module. Once the future state will be risky calculated by the risk state estimation module using current state information and the action outputted by the RL agent, the MCTS based safe policy search module will activate to guarantee a safer exploration by adding an additional reward for risk actions. We test the approach in several random overtake scenarios, resulting in faster convergence and safer behaviors compared to traditional reinforcement learning. [ABSTRACT FROM AUTHOR]

Interpretation of common-yet-challenging inter- action scenarios can benefit well-founded decisions for autonomous vehicles. Previous research achieved this using their prior knowledge of specific scenarios with predefined models, limiting their adaptive capabilities. This paper describes a Bayesian nonparametric approach that leverages continuous (i.e., Gaussian processes) and discrete (i.e., Dirichlet processes) stochastic processes to reveal underlying interaction patterns of the ego vehicle with other nearby vehicles. Our model relaxes dependency on the number of surrounding vehicles by developing an acceleration-sensitive velocity field based on Gaussian processes. The experiment results demonstrate that the velocity field can represent the spatial interactions between the ego vehicle and its surroundings. A discrete Bayesian nonparametric model, integrating Dirichlet processes and hidden Markov models, is developed to learn the interaction patterns over the temporal space by segmenting and clustering the sequential interaction data into interpretable granular patterns automatically. We then evaluate our approach in the highway discretionary lane-change scenarios using the highD dataset collected from real-world settings. Results demonstrate that our proposed Bayesian nonparametric approach provides an insight into the complicated discretionary lane-change interactions of the ego vehicle with multiple surrounding traffic participants based on the interpretable interaction patterns and their transition properties in temporal relationships. Our proposed approach sheds light on efficiently analyzing other kinds of multi-agent interactions, such as vehicle-pedestrian interactions. View associated demos via: https://chengyuan-zhang.github.io/Multivehicle-Interaction . [ABSTRACT FROM AUTHOR]

Although autonomous vehicle technology has evolved significantly in recent years, the navigation of self-driving vehicles in complex scenarios is still an open issue. One of the major challenges in these conditions is safe navigation on roads open to public traffic. The main issue is the interaction of the autonomous vehicle with regular traffic, as the behaviors and intentions of human-driven vehicles are hard to predict and understand. In this paper we propose a strategy to allow an autonomous vehicle to safely cross a multi-lane roundabout. Our approach uses a High-Definition (HD) map to predict at lane level the future situation, harnessing the concept of virtual instances of road users, which is a key concept in anticipating the situation in a roundabout that can be represented by a navigation graph with loops. This paper presents a methodology that uses intervals representing road occupancy by vehicles, with the road being widened to reflect uncertainties in localization. Our method safely avoids collisions and guarantees that no priority constraints are violated during the insertion maneuver. Moreover, the method does not provide an overly cautious insertion policy, i.e., an autonomous vehicle does not wait for a long time before the insertion. The performance of our strategy was evaluated using the SUMO simulation framework. To better evaluate the complexity of the simulation scenario, a highly interactive vehicle flow was generated using real dynamic traffic data from the INTERACTION dataset. We report real tests carried out with an experimental self-driving vehicle on a test circuit. Our results show that this approach is easy to integrate into an embedded system and that it allows roundabouts to be crossed with a high level of safety. [ABSTRACT FROM AUTHOR]

With recent advances in learning algorithms and hardware development, autonomous cars have shown promise when operating in structured environments under good driving conditions. However, for complex, cluttered, and unseen environments with high uncertainty, autonomous driving systems still frequently demonstrate erroneous or unexpected behaviors that could lead to catastrophic outcomes. Autonomous vehicles should ideally adapt to driving conditions; while this can be achieved through multiple routes, it would be beneficial as a first step to be able to characterize driveability in some quantified form. To this end, this paper aims to create a framework for investigating different factors that can impact driveability. Also, one of the main mechanisms to adapt autonomous driving systems to any driving condition is to be able to learn and generalize from representative scenarios. The machine learning algorithms that currently do so learn predominantly in a supervised manner and consequently need sufficient data for robust and efficient learning. Therefore, we also perform a comparative overview of 54 public driving datasets that enable learning and publish this dataset index at https://sites.google.com/view/driveability-survey-datasets. Specifically, we categorize the datasets according to the use cases and highlight the datasets that capture the complicated and hazardous driving conditions, which can be better used for training robust driving models. Furthermore, by discussions of what driving scenarios are not covered by the existing public datasets and what driveability factors need more investigation and data acquisition, this paper aims to encourage both targeted dataset collection and the proposal of novel driveability metrics that enhance the robustness of autonomous cars in adverse environments. [ABSTRACT FROM AUTHOR]

The path-following problem for autonomous vehicles is investigated in this paper. The desired vehicle heading is commonly chosen as the tangent direction on the desired path. This paper points out that the traditional definition of the desired heading may deteriorate the path-following performance, particularly when the vehicle is tracking a path with large curvature. That is because the sideslip angle control and the yaw rate control are conflicting in the presence of sliding effects, and the sideslip angle does not equal to zero when the vehicle is tracking a curve path. This paper further provides an amendment to the definition of the desired heading, which realizes a more accurate path-following maneuver. In the controller design phase, backstepping is used to generate the required yaw rate, and an LQR controller is adopted to obtain the optimal active front steering input. The CarSim–Simulink joint simulation verifies the reasonability of the amendment to the desired heading. [ABSTRACT FROM PUBLISHER]

Environment analysis of automatic vehicles needs the detection from 3-D point cloud information. This paper addresses this task when only partial scanning data are available. Our method develops the detection capabilities of autonomous vehicles equipped with 3-D range sensors for navigation purposes. In industrial practice, the safety scanners of automated guided vehicles (AGVs) and a localization technology provide an additional possibility to gain 3-D point clouds from planar contour points or low vertical resolution. Based on this data and a suitable evaluation algorithm, intelligence of vehicles can be significantly increased without the need for installation of additional sensors. In this paper, we propose a solution for an obstacle categorization problem for partial point clouds without shape modeling. The approach is tested for a known database, as well as for real-life scenarios. In case of AGVs, real-time run is provided by on-board computers of usual complexity. [ABSTRACT FROM AUTHOR]

The second edition of the grand cooperative driving challenge (GCDC2016) was held in The Netherlands in May 2016. Ten international teams participated in the two competition scenarios designed for GCDC2016: platoon merging and intersection. This paper describes the design and development of DRIVERTIVE, a DRIVERless cooperaTIve VEhicle, which aims to advance cooperative automation. The purpose of this paper is to give a general overview of the different designs used to adapt a factory vehicle, with no access to low-level control systems, into a fully-automated cooperative vehicle fit to compete in GCDC2016. The approach taken was pragmatic: different pre-existing techniques for control, state estimation, data fusion, communication, and data degradation were combined and experimentally validated in real-world scenarios, together with other vehicles with different implementations. Our main conclusion is that cooperative autonomous driving is feasible among very different implementations of the communication protocols and using completely different autonomous vehicles. [ABSTRACT FROM PUBLISHER]

Drivers involved in intersection collisions are at high risk of serious or fatal injury. Intersection advanced driver assistance systems (I-ADAS) are emerging active safety systems designed to help drivers safely traverse intersections. The effectiveness of I-ADAS is expected to be greatly dependent on pre-crash vehicle acceleration during intersection traversals. The objective of this paper was to develop pre-crash acceleration models for non-turning drivers involved in straight crossing path crashes and left-turning drivers in left turn across path opposite direction and lateral direction crashes. This paper used 348 event data recorder pre-crash records taken from crashes investigated as part of the National Automotive Sampling System/Crashworthiness Data System. The acceleration models generated from this pre-crash data were evaluated using a leave-one-out cross-validation procedure. Previously developed non-crash models from the literature were compared with the pre-crash models. Our hypothesis was that drivers involved in crashes would accelerate more aggressively than the “typical” driving population. This result suggests that drivers in pre-crash scenarios tend to accelerate more aggressively than drivers in normal scenarios (p<0.001). This has important implications for the design of I-ADAS. Specifically, higher acceleration results in less available time for I-ADAS to detect and respond to an imminent collision. [ABSTRACT FROM PUBLISHER]

Knowledge of the ego-vehicle’s motion state is essential for assessing the collision risk in advanced driver assistance systems or autonomous driving. Vision-based methods for estimating the ego-motion of vehicle, i.e., visual odometry, face a number of challenges in uncontrolled realistic urban environments. Existing solutions fail to achieve a good tradeoff between high accuracy and low computational complexity. In this paper, a framework for ego-motion estimation that integrates runtime-efficient strategies with robust techniques at various core stages in visual odometry is proposed. First, a pruning method is employed to reduce the computational complexity of Kanade–Lucas–Tomasi (KLT) feature detection without compromising on the quality of the features. Next, three strategies, i.e., smooth motion constraint, adaptive integration window technique, and automatic tracking failure detection scheme, are introduced into the conventional KLT tracker to facilitate generation of feature correspondences in a robust and runtime efficient way. Finally, an early termination condition for the random sample consensus (RANSAC) algorithm is integrated with the Gauss–Newton optimization scheme to enable rapid convergence of the motion estimation process while achieving robustness. Experimental results based on the KITTI odometry data set show that the proposed technique outperforms the state-of-the-art visual odometry methods by producing more accurate ego-motion estimation in notably lesser amount of time. [ABSTRACT FROM PUBLISHER]

Parking control problems remain to be fully solved for autonomous vehicles. Existing approaches usually first design a reference parking trajectory that does not exactly match vehicle dynamic constraints and then apply certain online negative feedback control to make the vehicle roughly track this reference trajectory. In this paper, we propose a novel trajectory planning method that directly links the actual parking trajectories and the steering actions to find the best parking trajectory. Tests show that this new approach has high reliability and less computation cost. Moreover, we also discuss how to counter with trajectory planning errors that are caused by model uncertainty in this paper. We show that an appropriate combination of feedforward trajectory planning and online feedback control can solve such problems. [ABSTRACT FROM PUBLISHER]

Motion planning is one of the most significant part in autonomous driving. Learning-based motion planning methods attract many researchers’ attention due to the abilities of learning from the environment and directly making decisions from the perception. The deep Q-network, as a popular reinforcement learning method, has achieved great progress in autonomous driving, but these methods seldom use the global path information to handle the issue of directional planning such as making a turning at an intersection since the agent usually learns driving strategies only by the designed reward function, which is difficult to adapt to the driving scenarios of urban roads. Moreover, different motion commands such as the steering wheel and accelerator are associated with each other from classic Q-networks, which easily leads to an unstable prediction of the motion commands since they are independently controlled in a practical driving system. In this paper, a conditional deep Q-network for directional planning is proposed and applied in end-to-end autonomous driving, where the global path is used to guide the vehicle to drive from the origination to the destination. To handle the dependency of different motion commands in Q-networks, we take use of the idea of fuzzy control and develop a defuzzification method to improve the stability of predicting the values of different motion commands. We conduct comprehensive experiments in the CARLA simulator and compare our method with the state-of-the-art methods. Experimental results demonstrate the proposed method achieves better learning performance and driving stability performance than other methods. [ABSTRACT FROM AUTHOR]

To autonomously navigate in traffic roads, an Autonomous Vehicle must take into account perception information, as well as the topological and geometric structure of the environment it is inserted in. Specifically in urban scenarios, the vehicle has to plan its path across intersections, roundabouts and perform lane changes to obey traffic rules. In addition, the planning algorithm must also consider the kinematics constraints of the vehicle and comfort parameters to the passengers. This paper proposes a road network model based on clothoids, which embraces the geometric and topological representation of the environment in a compact data structure. Piecewise linear continuous-curvature paths composed of clothoids, circular arcs, and straight lines are used for this purpose. The proposed approaches are evaluated in an urban scenario composed of curved and straight roads with single and double lanes, roundabouts, and intersections. As a result, a navigation architecture for Autonomous Vehicles was developed using the model, including global planning with continuous-curvature paths. [ABSTRACT FROM AUTHOR]

A crucial component of an autonomous vehicle (AV) is the artificial intelligence (AI) is able to drive towards a desired destination. Today, there are different paradigms addressing the development of AI drivers. On the one hand, we find modular pipelines, which divide the driving task into sub-tasks such as perception and maneuver planning and control. On the other hand, we find end-to-end driving approaches that try to learn a direct mapping from input raw sensor data to vehicle control signals. The later are relatively less studied, but are gaining popularity since they are less demanding in terms of sensor data annotation. This paper focuses on end-to-end autonomous driving. So far, most proposals relying on this paradigm assume RGB images as input sensor data. However, AVs will not be equipped only with cameras, but also with active sensors providing accurate depth information (e.g., LiDARs). Accordingly, this paper analyses whether combining RGB and depth modalities, i.e. using RGBD data, produces better end-to-end AI drivers than relying on a single modality. We consider multimodality based on early, mid and late fusion schemes, both in multisensory and single-sensor (monocular depth estimation) settings. Using the CARLA simulator and conditional imitation learning (CIL), we show how, indeed, early fusion multimodality outperforms single-modality. [ABSTRACT FROM AUTHOR]