Intelligent Autonomous Vehicles Research Articles

Transitioning from AGVs to AIVs in integrated job shop scheduling with transportation tasks: a multi-agent simulator for comparative analysis

Optimizing job shop scheduling in modern factories demands flexibility and adaptability to handle unexpected events and Unmanned Ground Vehicles (UGVs) limitations. This paper addresses these challenges by introducing a novel multi-agent simulator for the Job Shop Scheduling Problem (JSSP) with UGVs handling transportation tasks. The simulator, designed with Netlogo, incorporates real-world constraints, such as collision avoidance, UGV fleet size, and battery limitations, often overlooked in prior studies. By comparing the two categories of UGVs, namely, Autonomous Guided Vehicles (AGVs) and Autonomous Intelligent Vehicles (AIVs), under different scheduling methods (static vs dynamic), we evaluate their performance in constrained manufacturing environments. Our findings highlight the superior performance of AIVs in terms of overall makespan and resilience to additional constraints. Among other results, we found that schedules with a fleet of 2 AIVs produced similar makespans than schedules with a fleet of 4 AGVs. Consequently, the simulator and the conducted study provide valuable insights for optimizing JSSP within constrained environments and making informed decisions regarding AIVs adoption.

How does vortex dynamics help undulating bodies spread odor?

In this paper, we examine the coupling between odor dynamics and vortex dynamics around undulating bodies, with a focus on bio-inspired propulsion mechanisms. Utilizing computational fluid dynamics simulations with an in-house immersed boundary method solver, we investigate how different waveform patterns, specifically carangiform and anguilliform, influence the dispersion of chemical cues in both water and air environments. Our findings reveal that vortex dynamics significantly impact the overall trajectory of odor spots, although the alignment between odor spots and coherent flow structures is not always precise. We also evaluate the relative contributions of diffusion and convection in odor transport, showing that convection dominates in water, driven by higher Schmidt numbers, while diffusion plays a more prominent role alongside convection in air. Additionally, the anguilliform waveform generally produces stronger and farther-reaching chemical cues compared to carangiform swimmers. The critical roles of Strouhal number and Reynolds number in determining the efficiency of odor dispersion are also explained, offering insights that could enhance the design of more efficient, adaptive, and intelligent autonomous underwater vehicles by integrating sensory and hydrodynamic principles inspired by fish locomotion.

Deep Reinforcement Learning for Autonomous Driving Systems

Autonomous driving systems (ADS) are poised to revolutionize the future of transportation, promising increased safety, efficiency, and convenience. Deep Reinforcement Learning (DRL) has emerged as a powerful approach to solving complex decision-making tasks in dynamic environments, making it a promising candidate for the development of intelligent autonomous vehicles. This paper explores the application of DRL techniques in autonomous driving, focusing on the integration of perception, planning, and control. We review state-of-the-art DRL algorithms, including Deep Q-Networks (DQN), Proximal Policy Optimization (PPO), and Soft Actor-Critic (SAC), and examine their roles in enabling end-to-end learning for driving policies. Furthermore, we discuss the challenges inherent in deploying DRL in real-world autonomous driving scenarios, including sample inefficiency, safety constraints, and the sim-to-real gap. Finally, the paper presents case studies and experimental results that highlight the potential of DRL to improve autonomous vehicle performance in complex environments, while identifying future research directions to address open problems in the field.

Investigating the Impacts of Autonomous Vehicles on the Efficiency of Road Network and Traffic Demand: A Case Study of Qingdao, China.

Rapid urbanization has led to the development of intelligent transport in China. As active safety technology evolves, the integration of autonomous active safety systems is receiving increasing attention to enable the transition from functional to all-weather intelligent driving. In this process of transformation, the goal of automobile development becomes clear: autonomous vehicles. According to the Report on Development Forecast and Strategic Investment Planning Analysis of China's autonomous vehicle industry, at present, the development scale of China's intelligent autonomous vehicles has exceeded market expectations. Considering limited research on utilizing autonomous vehicles to meet the needs of urban transportation (transporting passengers), this study investigates how autonomous vehicles affect traffic demand in specific areas, using traffic modeling. It examines how different penetration rates of autonomous vehicles in various scenarios impact the efficiency of road networks with constant traffic demand. In addition, this study also predicts future changes in commuter traffic demand in selected regions using a constructed NL model. The results aim to simulate the delivery of autonomous vehicles to meet the transportation needs of the region.

Optimized Design of EdgeBoard Intelligent Vehicle Based on PP-YOLOE.

Advances in deep learning and computer vision have overcome many challenges inherent in the field of autonomous intelligent vehicles. To improve the detection accuracy and efficiency of EdgeBoard intelligent vehicles, we proposed an optimized design of EdgeBoard based on our PP-YOLOE+ model. This model innovatively introduces a composite backbone network, incorporating deep residual networks, feature pyramid networks, and RepResBlock structures to enrich environmental perception capabilities through the advanced analysis of sensor data. The incorporation of an efficient task-aligned head (ET-head) in the PP-YOLOE+ framework marks a pivotal innovation for precise interpretation of sensor information, addressing the interplay between classification and localization tasks with high effectiveness. Subsequent refinement of target regions by detection head units significantly sharpens the system's ability to navigate and adapt to diverse driving scenarios. Our innovative hardware design, featuring a custom-designed mainboard and drive board, is specifically tailored to enhance the computational speed and data processing capabilities of intelligent vehicles. Furthermore, the optimization of our Pos-PID control algorithm allows the system to dynamically adjust to complex driving scenarios, significantly enhancing vehicle safety and reliability. Besides, our methodology leverages the latest technologies in edge computing and dynamic label assignment, enhancing intelligent vehicles' operations through seamless sensor integration. Our custom dataset, specifically designed for this study, includes 4777 images captured by intelligent vehicles under a variety of environmental and lighting conditions. The dataset features diverse scenarios and objects pertinent to autonomous driving, such as pedestrian crossings and traffic signs, ensuring a comprehensive evaluation of the model's performance. We conducted extensive testing of our model on this dataset to thoroughly assess sensor performance. Evaluated against metrics including accuracy, error rate, precision, recall, mean average precision (mAP), and F1-score, our findings reveal that the model achieves a remarkable accuracy rate of 99.113%, an mAP of 54.9%, and a real-time detection frame rate of 192 FPS, all within a compact parameter footprint of just 81 MB. These results demonstrate the superior capability of our PP-YOLOE+ model to integrate sensor data, achieving an optimal balance between detection accuracy and computational speed compared with existing algorithms.

An All-Terrain Intelligent Autonomous Vehicle with Sensor Fusion Based Navigation Capabilities

An automated defense robot building is planned, that has a laser gun attached, which is utilized for pointing laser rays to destroy the target object. One of the most important things about these robots is that they have the capability to perform missions remotely in the field, without any actual danger to human lives. In the proposed system, a robot is developed that is controlled through Microcontroller. For the aiming purpose and to view the road and the surroundings in which the robot is travelling, wireless cameras are installed. Radio frequency can be used to control the robot. These defense robots used in military are usually employed with the integrated system including gripper, cameras and sensors. This is specially designed for defense robotic system to save human life and protect the country from enemies. In the existing systems, personal computer using ZigBee protocol is used to monitor the robot.

Blockage Effects of Road Bridge on mmWave Channels for Intelligent Autonomous Vehicles

Vehicular communication and sensing technologies are key to enabling 6G Intelligent Autonomous Transportation Systems (IATS). With the introduction of massive sensors and artificial intelligence (AI) fusion applications, IATS is needed to support data transmission rates up to 10 Gb/s. Millimeter-wave (mmWave) technology has attracted extensive attention owing to abundant spectrum resources, which can support the timely transmission of massive data. However, performance degradation of mmWave due to signal blockage has become one of the critical technical challenges. Road bridges as one of the common obstacles in urban scenarios, which has severe blockage effects on communication links. Therefore, this paper comprehensively studies the impact of road bridge blockage effects on mmWave vehicle-to-infrastructure (V2I) links and proposes an empirical model that can accurately characterize the bridge blockage effect. First, we use a self-developed mmWave channel sounder to carry out channel measurements on typical urban roads. Measurement results indicate that a maximum extra propagation loss of up to 23 dB is caused by road bridges. In addition, to address the deficiencies of existing propagation prediction models, the Single Road Bridge (SRB) model is proposed in this work. This model reveals for the first time the extra propagation loss caused by the road bridge to the channel. Compared with existing models, the SRB model can make the mean absolute error (MAE) and root mean square error (RMSE) within 5 dB. The proposed SRB model is of great value for accurately simulating real-world road bridge blockage events when designing future IATS.

Digital supply chain: literature review of seven related technologies

This paper systematically reviews literature related with digital supply chains (DSC) and investigates the application status and development trend of different digital technologies in supply chain management. The review is conducted from the perspective of seven key digital supply chain technologies, i.e. Internet of Things (IoT) & Radio Frequency Identification (RFID), 5th Generation Mobile Communication Technology (5G), 3D Printing, Big data (BD), Blockchain, Digital Twins (DT), and Intelligent autonomous vehicles (IAVs). It highlights the main limitations and opportunities of the various DSC technologies, provides an overview of prior studies, and identifies knowledge gaps by outlining the advantages, weaknesses and restrictions of individual technology. The paper also aims at providing a development framework as a roadmap for the match of different digital technologies with different strategic goals.

Cluster‐based context‐aware route service management for smart intelligent autonomous vehicles with industrial transport system

SummaryIndustrial transport system refers to the movement of goods, raw materials, and finished products within and between industrial facilities such as factories, warehouses, and distribution centers. It includes a variety of modes of transport such as trucks, forklifts, conveyor belts, and automated guided vehicles (AGVs). Context‐aware route service management refers to system that can dynamically manage and optimize the routes taken by vehicles in transportation system based on contextual information. Our proposed framework for industrial transport systems and smart intelligent autonomous vehicles utilizes cluster‐based context‐aware routing (CCAR), which includes several significant contributions. Firstly, modified elephant herding optimization (MEHO) algorithm is used to efficiently group autonomous vehicles for routing. Secondly, an improved gravitational search (IGS) algorithm is used for cluster head (CH) selection responsible for transferring context information between vehicles (V2V). Thirdly, a deep hybrid multi‐graph neural network (DHMG‐NN) is designed for optimal neighboring roadside unit (RSU) node selection through different design constraints to ensure data dissemination between vehicles and infrastructure (V2I). Finally, we validate the effectiveness of our CCAR framework using various simulation scenarios. Our framework achieves a reduction of up to 78% in computation time and an improvement of up to 25% in customer satisfaction ratio and 15% higher than the overall performance ratio compared to the existing frameworks.

Identify, Estimate and Bound the Uncertainty of Reinforcement Learning for Autonomous Driving

Deep reinforcement learning (DRL) has emerged as a promising approach for developing more intelligent autonomous vehicles (AVs). A typical DRL application on AVs is to train a neural network-based driving policy. However, the black-box nature of neural networks can result in unpredictable decision failures, making such AVs unreliable. To this end, this work proposes a method to identify and protect unreliable decisions of a DRL driving policy. The basic idea is to estimate and constrain the policy’s performance uncertainty, which quantifies potential performance drop due to insufficient training data or network fitting errors. By constraining the uncertainty, the DRL model’s performance is always greater than that of a baseline policy. The uncertainty caused by insufficient data is estimated by the bootstrapped method. Then, the uncertainty caused by the network fitting error is estimated using an ensemble network. Finally, a baseline policy is added as the performance lower bound to avoid potential decision failures. The overall framework is called uncertainty-bound reinforcement learning (UBRL). The proposed UBRL is evaluated on DRL policies with different amounts of training data, taking an unprotected left-turn driving case as an example. The result shows that the UBRL method can identify potentially unreliable decisions of DRL policy. The UBRL guarantees to outperform baseline policy even when the DRL policy is not well-trained and has high uncertainty. Meanwhile, the performance of UBRL improves with more training data. Such a method is valuable for the DRL application on real-road driving and provides a metric to evaluate a DRL policy.

Joint Optimization of Depth and Ego-Motion for Intelligent Autonomous Vehicles

The three-dimensional (3D) perception of autonomous vehicles is crucial for localization and analysis of the driving environment, while it involves massive computing resources for deep learning, which can't be provided by vehicle-mounted devices. This requires the use of seamless, reliable, and efficient massive connections provided by the 6G network for computing in the cloud. In this paper, we propose a novel deep learning framework with 6G enabled transport system for joint optimization of depth and ego-motion estimation, which is an important task in 3D perception for autonomous driving. A novel loss based on feature map and quadtree is proposed, which uses feature value loss with quadtree coding instead of photometric loss to merge the feature information at the texture-less region. Besides, we also propose a novel multi-level V-shaped residual network to estimate the depths of the image, which combines the advantages of V-shaped network and residual network, and solves the problem of poor feature extraction results that may be caused by the simple fusion of low-level and high-level features. Lastly, to alleviate the influence of image noise on pose estimation, we propose a number of parallel sub-networks that use RGB image and its feature map as the input of the network. Experimental results show that our method significantly improves the quality of the depth map and the localization accuracy and achieves the state-of-the-art performance.

An efficient approach for traffic sign detection, classification, and localization applied for autonomous intelligent vehicles

In this paper, an efficient approach for traffic sign detection, classification, and localization is introduced. An integration of the tiling technique with the RetinaFace model and the MobileNetV1-SSD was proposed for traffic detection and classification. The combination of the extraction of the region of interest with a depth estimation model, namely the AANet+, was introduced for the traffic sign localization task. All models were developed based on the transfer learning technique and existing datasets, including the Zalo and ApolloScape datasets. The accuracy and the computational efficiency of the approach are evaluated. Experimental results show that the novel traffic sign detection and classification method outperform the existing ones with an average precision of 77.2%. Moreover, the computing performance achieved is 5 FPS on Jetson Nano and 50 FPS on Jetson Xavier. For the traffic sign localization, the relative error can be reduced to 3.78%, and the computing time is 1.965 s/pair on Jetson Nano and 0.128 s/pair on Jetson Xavier, while the existing method has lower accuracy and is more time-consuming.

Vision-Based Ingenious Lane Departure Warning System for Autonomous Vehicles

Lane detection is necessary for developing intelligent Autonomous Vehicles (AVs). Using vision-based lane detection is more cost-effective, requiring less operational power. Images captured by the moving vehicle include varying brightness, blur, and occlusion caused due to diverse locations. We propose a Vision-based Ingenious Lane Departure Warning System (VILDS) for AV to address these challenges. The Generative Adversarial Networks (GAN) of the VILDS choose the most precise features to create images that are identical to the original but have better clarity. The system also uses Long Short-Term Memory (LSTM) to learn the average behavior of the samples to forecast lanes based on a live feed of processed images, which predicts incomplete lanes and increases the reliability of the AV’s trajectory. Further, we devise a strategy to improve the Lane Departure Warning System (LDWS) by determining the angle and direction of deviation to predict the AV’s Lane crossover. An extensive evaluation of the proposed VILDS system demonstrated the effective working of the lane detection and departure warning system modules with an accuracy of 98.2% and 96.5%, respectively.

Scheduling AIV transporter using simulation-based supervised learning: A case study on a dynamic job-shop with three workstations

Dynamic job shop scheduling consists of scheduling jobs dynamically with different routing on a set of machines. A feasible and quick solution can be computed using heuristics. One well-known heuristic in job shop problems is selecting the priority dispatching rule (such as giving priority to the job with the Shortest Processing Time called SPT). Nowadays, with the application of industry 4.0 technologies such as sensors, Intelligent robots, etc., workshop data are more accessible and can be exploited to find the appropriate dispatching rule depending on the state of the shop. This work proposes a data-driven methodology for scheduling an AIV (Autonomous Intelligent Vehicle) that supplies three workstations. Our approach is based on data collected from an Arena simulation model fed to a supervised learning algorithm. This one helps identify one among five dispatching rules for each scheduling decision.

Intelligent autonomous underwater vehicle mobility with energy efficient routing in sensor networks

Intelligent autonomous underwater vehicle mobility with energy efficient routing in sensor networks

Reliable and Seamless Communications of Networked IAV in Container Terminal Using UAV Technology

Abstract Managing container loading and unloading operations at a container terminal using Intelligent Autonomous Vehicle (IAV) is challenging, especially at intersections in the yard, which are often inevitable. For ensuring efficient and accident-free management of these intersections, the IAV must cooperate by exchanging messages. Due to signal obstruction at these intersections, indirect communication is established through an additional relay node to ensure reliable communication. This paper proposes distributed approach using Unmanned Aerial Vehicle (UAV) connectivity to avoid collision and deadlock between IAV at intersections. Due to the obstacles formed by stacked containers blocking the radio transmissions, the proposed algorithm automatically switches the ongoing communication through the UAV to ensure successful communication at intersections. Thus, the idea of introducing UAV as communication relay in a container terminal is an interesting solution that we have adopted in this work. Simulation techniques are used to evaluate our proposal. The obtained results confirm that our UAV-based approach ensures reliable communication and automated intersections management in the yard while further ensuring the safety of IAV traffic.

Development of Real-Time Traffic-Object and Traffic-Sign Detection Models Applied for Autonomous Intelligent Vehicles

Technologies for detecting traffic objects are an essential requirement for any applications in autonomous intelligent vehicles. In this work, models for detecting traffic objects were developed. Based on the existing datasets and the pre-trained models, fine-tuning techniques were applied to achieve trained models with higher accuracies even for the very challenging test data. The traffic object detection was developed based on the pre-trained YOLOv5s model. Two approaches were introduced for the traffic sign detection task. The so-called tiling technique incorporated with the YOLOv5s model was exploited in the first approach. In the second approach, a combination of the RetinaFace model for the localization and the MobileNetV1-SSD for the classification was employed. The experimental results show that all developed models release a very high rate of accuracy with a maximum AP50 of 75.9% for object detection and mAP50 of 64.2% for sign detection. Models developed via the second approach have twofold advantages in terms of accuracy and computational efficiency, which allows to deploy practical applications.

Investigation of New Prosumer Peer-to-Peer Decisions into Energy Community Integration

The search for advanced solutions for Prosumers and their integration is one of the critical directions of forming the Energy Community. Future challenges to electric grids and their solutions and other subject areas in the energy sector include such related issues as Intelligent Autonomous Vehicles, Monitoring, Protection, and Control in the Post-Cyber Attacks for the formation of integration components of progressive ecosystems. The report provides a brief overview from the perspective of the development of the Proactive Control and the study of the applicability of the Data Environment Analysis technology.

Safety driven intelligent autonomous vehicle for smart cities using IoT

Purpose Autonomous vehicles rely on IoT-based technologies to take numerous decisions in real-time situations. However, added information from the sensor readings will burden the system and cause the sensors to produce inaccurate readings. To overcome these issues, this paper aims to focus on communication between sensors and autonomous vehicles for better decision-making in real-time. The system has unique features to detect the upcoming and ongoing vehicles automatically without intervention of humans in the system. It also predicts the type of vehicle and intimates the driver. Design/methodology/approach The system is designed using the ATmega 328 P and ESP 8266 chip. Information from ultrasonic and infrared sensors are analyzed and updated in the cloud server. The user can access all these real-time data at any point of time. The stored information in cloud servers is used for integrating artificial intelligence into the system. Findings The real-time sensor information is used to predict the surrounding environment and the system responds to the user according to the situation. Practical implications The system is implemented on embedded platform with IoT technology. The sensor information is updated to the cloud using the Blynk application for the user in real time. Originality/value The system is proposed for smart cities with IoT technology where the user and the system are aware of the surrounding environment. The system is mainly concerned with the accuracy of sensors and the distance between the vehicles in real-time environment.

Mobile Robot Obstacle Avoidance Based on Neural Network with a Standardization Technique

Reactive algorithm in an unknown environment is very useful to deal with dynamic obstacles that may change unexpectantly and quickly because the workspace is dynamic in real-life applications, and this work is focusing on the dynamic and unknown environment by online updating data in each step toward a specific goal; sensing and avoiding the obstacles coming across its way toward the target by training to take the corrective action for every possible offset is one of the most challenging problems in the field of robotics. This problem is solved by proposing an Artificial Intelligence System (AIS), which works on the behaviour of Intelligent Autonomous Vehicles (IAVs) like humans in recognition, learning, decision making, and action. First, the use of the AIS and some navigation methods based on Artificial Neural Networks (ANNs) to training datasets provided high Mean Square Error (MSE) from training on MATLAB Simulink tool. Standardization techniques were used to improve the performance of results from the training network on MATLAB Simulink. When it comes to knowledge-based systems, ANNs can be well adapted in an appropriate form. The adaption is related to the learning capacity since the network can consider and respond to new constraints and data related to the external environment.