Autonomous Drone Navigation Research Articles

Neural network model for autonomous navigation of a water drone

Water drones have significant potential for use in environmental monitoring, search and rescue operations, and marine infrastructure inspection, but the specific conditions of the water environment make it difficult to implement stable autonomous navigation. The object of research presented in this paper is the machine learning process for autonomous navigation of a water drone model in a simulated water environment. The purpose of the study is to implement a neural network model for autonomous navigation of a water drone using a reinforcement learning method that provides improved obstacle avoidance and adaptation to water currents. To achieve this purpose, a new neural network model for autonomous drone navigation in the water environment based on the reinforcement learning method is proposed, which differs from the existing ones in that it uses an improved drone control algorithm that takes into account the speed and direction of the water current, which makes it possible to stabilize the process of generating neural network coefficients. To ensure an effective learning process and optimization of the model, a simulation training environment was developed using the USVSim simulator, which contains various factors that interfere with the drone's movement, such as water current and the presence of other objects. The water drone, acting as an agent, gradually learns to choose the most effective actions to maximize positive rewards through trial and error, interacting with the environment and adapting to changing conditions. This process takes place through the use of a Deep Q-Network: the drone provides the value of its current state to a deep neural network; the neural network processes the data, predicts the value of the most effective action, and gives it to the agent. The current state of the drone is information in the form of a set of sensor readings measuring the distance to the nearest obstacles, drone’s heading and current distance to goal. The value of the effective action received from the neural network is converted into a command for the rudder that the drone can understand. The value of the drone's thruster power is calculated by separate formulas using trigonometric functions. The results of the study showed that the use of the proposed model allows the drone to make decisions in a dynamic water environment when rapid adaptation to changes is required. The model successfully adapted its strategy based on feedback from the environment, so it can be concluded that the implemented model shows significant potential for further research and applications in the field of autonomous water drones, especially in changing and unpredictable environments.

Precision without GPS: Multi-Sensor Fusion for Autonomous Drone Navigation in Complex Environments

The rapid evolution of drone technology has expanded its applications across variodus domains, including delivery services, environmental monitoring, and search and rescue operations. However, many of these applications face significant challenges in GPS-denied environments, such as dense urban areas and heavily forested regions, where traditional navigation methods falter. This paper presents a novel multi-sensor fusion algorithm designed to enhance the localization accuracy of autonomous drones without reliance on GPS. By integrating data from an Inertial Measurement Unit (IMU), LiDAR, and visual sensors, the proposed approach effectively compensates for the limitations of individual sensors, enabling robust navigation in complex environments. Experimental results demonstrate that the algorithm achieves an average localization accuracy of 1.2 meters in urban areas and 1.5 meters in forested settings, showcasing its resilience against sensor noise and environmental challenges. The implementation of loop closure techniques further improves long-term navigation accuracy, making it suitable for prolonged missions. This research contributes to the growing body of knowledge in autonomous drone navigation and offers significant implications for enhancing the operational capabilities of drones in real-world scenarios. Future work will focus on integrating additional sensors, exploring machine learning techniques for adaptive fusion, and conducting extensive field trials to validate the system's performance in dynamic environments..

Autonomous Drone Navigation Using Computer Vision

Abstract: Autonomous drone navigation with computer vision is an innovative technology that allows drones to move through intricate surroundings without needing human control. This system uses live visual information to identify objects, steer clear of obstacles, and determine routes, improving operational safety and precision. The project's main objective is to create a visionbased navigation system that combines object detection and obstacle avoidance algorithms through deep learning methods like YOLO (You Only Look Once), alongside real-time sensor fusion. The drone uses computer vision algorithms to process aerial images and automatically changes its flight path to prevent crashes. A personalized dataset of aerial images is generated and utilized for improving object detection skills in various environments. In order to guarantee practicality in real- world situations, the system undergoes validation through simulations and field tests on different terrains, focusing on its resilience in changing environments. Improvements in navigation accuracy and obstacle detection are accomplished by implementing adaptive pathplanning and integrating multiple sensors, guaranteeing the drone's efficient operation in real-life situations. The goal of this method is to enhance the drone's ability to make decisions, minimize human mistakes, and expand its range of potential uses in areas like surveillance, agriculture, and disaster response

Testing a Vision-Based Autonomous Drone Navigation Model in a Forest Environment

Drones play a pivotal role in various industries of Industry 4.0. For achieving the application of drones in a dynamic environment, finding a clear path for their autonomous flight requires more research. This paper addresses the problem of finding a navigation path for an autonomous drone based on visual scene information. A deep learning-based object detection approach can localize obstacles detected in a scene. Considering this approach, we propose a solution framework that includes masking with a color-based segmentation method to identify an empty area where the drone can fly. The scene is described using segmented regions and localization points. The proposed approach can be used to remotely guide drones in dynamic environments that have poor coverage from global positioning systems. The simulation results show that the proposed framework with object detection and the proposed masking technique support drone navigation in a dynamic environment based only on the visual input from the front field of view.

End-to-End Nano-Drone Obstacle Avoidance for Indoor Exploration

Autonomous navigation of drones using computer vision has achieved promising performance. Nano-sized drones based on edge computing platforms are lightweight, flexible, and cheap; thus, they are suitable for exploring narrow spaces. However, due to their extremely limited computing power and storage, vision algorithms designed for high-performance GPU platforms cannot be used for nano-drones. To address this issue, this paper presents a lightweight CNN depth estimation network deployed on nano-drones for obstacle avoidance. Inspired by knowledge distillation (KD), a Channel-Aware Distillation Transformer (CADiT) is proposed to facilitate the small network to learn knowledge from a larger network. The proposed method is validated on the KITTI dataset and tested on a Crazyflie nano-drone with an ultra-low power microprocessor GAP8. This paper also implements a communication pipe so that the collected images can be streamed to a laptop through the on-board Wi-Fi module in real-time, enabling an offline reconstruction of the environment.

Multi-objective reinforcement learning for autonomous drone navigation in urban areas with wind zones

Drones can be used for tasks such as data collection and logistics in civil engineering. Current research on autonomous drones mainly focuses on planning a safe path and avoiding obstacles in a static environment. However, navigating a drone in complex environments like urban areas involves many dynamic constraints, such as building layout, winds, and signal coverage, which are interdependent. The wind factor is the most important among these environmental factors, which may cause a drone to lose control or even crash. This paper presents a multi-objective navigation reinforcement learning algorithm (MONRL) for the drone to navigate and avoid obstacles in an unknown environment when dynamic wind zones present, with only imagery data about the building layouts. Based on a deep reinforcement learning and memory architecture, the drone develops policies to prioritize navigation decisions, optimizing the path while minimizing negative impact of winds with only sparse sensor data, in our case, camera inputs. By leveraging the advantages of the proposed method in estimating the environmental factors from previous trials, no aerodynamic force sensors are needed for the drone to develop effective strategies to navigate to target while counteracting to milder winds and dodging aways from stronger winds. The proposed method was tested in a virtual environment, and a real model of New York City. The method is expected to contribute to new automation algorithms for urban aerial logistics and future automated civil infrastructure inspection.

Automatic correction of barometric altimeters using additional air temperature and humidity measurements

A GPS module is often combined with an atmospheric pressure sensor in onboard navigation systems. Thanks to this, using appropriate algorithms comparing the results obtained from GPS and from the barometric altimeter, it is possible to obtain much more accurate indications of the height above sea level. The authors analyzed the problem of correcting barometric altimeter errors caused by non-standard pressure and temperature conditions at sea level and the presence of water vapor in the air. The authors proposed methods of correcting the altimeter error using additional temperature and air humidity measurements. They also proposed a simple method for calibrating the altimeter prior to the mission to estimate the difference in pressure above the standard value. The paper discusses the influence of water vapor on the accuracy of the altimeter indications and the conditions for the applicability of the error correction method. The authors initially considered the possibility of using the method based on the concept of virtual temperature. However, this method was rejected because simulation studies showed that it did not give satisfactory results. Simulation tests were carried out in the MATLAB environment to assess the effectiveness of the proposed methods. The test results turned out to be satisfactory. They can be useful when designing low-cost navigation systems using simple sensors made in MEMS technology. The main envisaged application is in autonomous drone navigation systems.

Autonomous Drone Navigation

Autonomous drone navigation has evolved significantly as a result of the integration of cutting-edge computer vision and reinforcement learning algorithms. In this study, we propose a comprehensive framework that combines the adaptability of the TD3 neural network, the agility of the Ros2 platform, and the resilience of the YOLOv3 object detection model. Making use of YOLOv3's real-time object detection capabilities, our system proves to be highly adept at recognizing and reacting to changing environmental impediments, which guarantees improved UAV situational awareness. When combined with the TD3 neural network, the Ros2-based navigation framework allows for quick decision- making, which allows the drone to move independently toward predetermined destination points while effectively avoiding obstructions. Our work demonstrates the effectiveness of our suggested strategy through thorough experimentation in simulated situations, highlighting its potential for practical implementation in a variety of applications, from autonomous delivery services to aerial surveillance. The outcomes demonstrate the integrated system's flexibility and resilience, highlighting its capacity to successfully handle challenging navigational situations and guarantee secure and effective aerial operations. Our study adds to the field of autonomous drone navigation by highlighting the vital role that integrated computer vision and machine learning techniques play in enabling smart and flexible unmanned aerial vehicle operations. Keywords— Autonomous drone, YOLOv3, Object detection, Ros2, TD3, Reinforcement learning, Simulated environment, Aerial navigation, obstacle avoidance, Computer vision, Unmanned aerial vehicles (UAVs),Goal-point navigation, Aerial surveillance

AirLoop: A Simulation Framework for Testing of UAV Services

Sensor fusion is a critical aspect in autonomous drone navigation as several tasks, such as object detection and self-pose estimation, require combining information from heterogeneous sources. The performance of these solutions depends on several factors, such as the characteristics of the sensors and the environment, as well as the computing platforms, which can heavily impact their accuracy and response time. Carrying out such performance evaluations through real flight tests can be a resource-demanding, time-consuming, and, at times, risky process, which is why researchers often rely on simulation environments for testing and validating sensor fusion algorithms. The simulation environment should provide photorealistic environmental features, as well as a comprehensive set of sensors, in order to allow to test the most extensive set of sensor fusion algorithms. This paper presents AirLoop, an AirSim-based flight simulator for Hardware-in-the-Loop and Software-in-the-Loop algorithm testing and validation. AirLoop extends the sensor setup provided by AirSim with an FMCW RADAR sensor simulation, which has been evaluated based on the Infineon Technologies BGT60TR13C RADAR. Furthermore, this work provides several Software-in-the-Loop (SITL) and Hardware-in-the-Loop (HITL) demonstrations, including interfacing with the Pixhawk 2 flight controller and an extensive evaluation of the communication of the engine with the NVIDIA Jetson Nano, which has been evaluated in various use cases, including dataset creation, object detection, Path Planning, and Simultaneous Localization and Mapping (SLAM).

UGVs for Agri Spray with AI assisted Paddy Crop disease Identification

This research work involves data access and control of Unmanned Ground Vehicles (UGVs) which is designed for health monitoring of paddy crops as an exception from traditional monitoring system. The proposed scheme of autonomous navigation of drones is planned to be carried out as per the six milestones for disease detection and control in paddy crops. Drones serve as Unmanned Robotic Vehicles (URV) capable of performing desired tasks in unstructured, uncertain and potentially hostile environments and are remotely-operated without human intervention. URVs function completely as autonomous entities in diversified environments. Current UGVs adhere to different levels of automaticity. Typically the vehicle follows high level waypoints spaced for few hundred meters of distance to provide monitoring of agricultural fields and early detection of the various diseases that may occur in the paddy fields in a polyhouse. To increase the vehicle's abilities, tracking efficiency, obstacle avoidance, path planning or lead and follow up augmented control with Fuzzy Logic Controller (FLC) is incorporated. The distributed autonomous system for information gathering related to the paddy crops in polyhouse is enabled using different sensors, which is a data-intensive task. To increase the robustness of the system, fuzzy controllers are proposed to control the navigation of the proposed UGV in “All terrain conditions”. They are needed to offer problem specific heuristic control knowledge for the Inference Engine Design which occurs due to imprecision and uncertainty of the sensor readings. It also requires low computation time which favours the polyhouse situations. The navigation of UGV and the FLC action will in turn depend on “All terrain traversability” and “dead zone” monitoring. The proposed UGV model is capable of measuring the parameters associated with the paddy crops inside a polyhouse. The various diseases in paddy crops are False Smut (FS), Sheath Blight (SB), Rice Blast (RB), Leaf Scald (LS), Brown Spot (BS), Bacterial Leaf Blight (BLB) and Bakane (BE) which are detected using Convolutional Neural Network (CNN).

Exploring the Benefits of Reinforcement Learning for Autonomous Drone Navigation and Control

Drones are now used in a wide range of industries, including delivery services and agriculture. Notwithstanding, controlling robots in powerful conditions can be testing, particularly while performing complex assignments. Conventional strategies for drone mechanization depend on pre-customized directions, restricting their adaptability and versatility. Drones can learn from their interactions with their environment and improve their performance over time with the help of reinforcement learning (RL), which has emerged as a promising method for drone automation in recent years. This paper looks at how RL can be used to automate drones and how it can be used in different industries. In addition, the difficulties of RL-based drone automation and potential directions for future research are discussed in the paper.

Reinforcement Learning for Autonomous Drone Navigation

Drone navigation involves the process of controlling the movement and flight path of unmanned aerial vehicles (UAVs). It encompasses both the hardware and software systems that enable drones to navigate and maneuver autonomously or under the guidance of a human operator. The utility of drone navigation is vast and varied, making it a critical component in numerous industries and applications. Firstly, drone navigation plays a crucial role in aerial surveillance and reconnaissance. Drones equipped with advanced navigation systems can efficiently patrol large areas, monitor activities, and gather real-time data from various perspectives. This capability is particularly valuable in security and law enforcement operations, disaster response, and environmental monitoring, where access and visibility might be limited.Drone navigation is critical in aerial mapping and surveying. Drones outfitted with GPS and other positioning technologies may record photographs and collect data with pinpoint accuracy, allowing for the creation of comprehensive 3D maps, topographic models, and land surveys. This allows for more precise, quicker, and cost-effective data collecting for urban planning, agricultural, infrastructure inspection, and building projects. Furthermore, drone navigation is critical in the transportation and logistics industries. Autonomous drones can fly predetermined paths to rapidly and effectively transfer commodities, medical supplies, and other payloads to distant or inaccessible regions. This technology has the potential to transform last-mile distribution, especially in rural areas or during emergencies when regular transportation routes may be hampered.

Near-Ground Delivery Drones Path Planning Design Based on BOA-TSAR Algorithm

With the advancement of technology and the rise of the unmanned aerial vehicle industry, the use of drones has grown tremendously. For drones performing near-ground delivery missions, the problem of 3D space-based path planning is particularly important in the autonomous navigation of drones in complex spaces. Therefore, an improved butterfly optimization (BOA-TSAR) algorithm is proposed in this paper to achieve the autonomous pathfinding of drones in 3D space. First, this paper improves the randomness strategy of the initial population generation in the butterfly optimization algorithm (BOA) via the Tent chaotic mapping method, by means of the removal of the short-period property, which balances the equilibrium of the initial solutions generated by the BOA algorithm in the solution space. Secondly, this paper improves the shortcomings of the BOA algorithm in terms of slower convergence, lower accuracy, and the existence of local optimal stagnation when dealing with high-dimensional complex functions via adaptive nonlinear inertia weights, a simulated annealing strategy, and stochasticity mutation with global adaptive features. Finally, this paper proposes an initial population generation strategy, based on the 3D line of sight (LOS) detection method, to further reduce the generation of path interruption points while ensuring the diversity of feasible solutions generated by the BOA algorithm for paths. In this paper, we verify the superior performance of BOA-TSAR by means of simulation experiments. The simulation results show that BOA-TSAR is very competitive among swarm intelligence (SI) algorithms of the same type. At the same time, the BOA-TSAR algorithm achieves the optimal path length measure and smoothness measure in the path-planning experiment.

Designing obstacle’s map of an unknown place using autonomous drone navigation and web services

PurposeDrones are used in several purposes including examining areas, mapping surroundings and rescue mission operations. During these tasks, they could encounter compound surroundings having multiple obstacles, acute edges and deadlocks. The purpose of this paper is to propose an obstacle dodging technique required to move the drones autonomously and generate the obstacle's map of an unknown place dynamically.Design/methodology/approachTherefore, an obstacle dodging technique is essentially required to move autonomously. The automaton of drones requires complicated vision sensors and a high computing force. During this research, a methodology that uses two basic ultrasonic-oriented proximity sensors placed at the center of the drone and applies neural control using synaptic plasticity for dynamic obstacle avoidance is proposed. The two-neuron intermittent system has been established by neural control. The synaptic plasticity is used to find turning angles from different viewpoints with immediate remembrance, so it helps in decision-making for a drone. Hence, the automaton will be able to travel around and modify its angle of turning for escaping objects during the route in unknown surroundings with narrow junctions and dead ends. Furthermore, wherever an obstacle is detected during the route, the coordinate information is communicated using RESTful Web service to an android app and an obstacle map is generated according to the information sent by the drone. In this research, the drone is successfully designed and automated and an obstacle map using the V-REP simulation environment is generated.FindingsSimulation results show that the drone effectively moves and turns around the obstacles and the experiment of using web services with the drone is also successful in generating the obstacle's map dynamically.Originality/valueThe obstacle map generated by autonomous drone is useful in many applications such as examining fields, mapping surroundings and rescue mission operations.

Assessment of Android Network Positioning as an Alternative Source of Navigation for Drone Operations

Applications of drones have increased significantly in the past decade for both indoor and outdoor operations. In order to assist autonomous drone navigation, there are numerous sensors installed onboard the vehicles. These include Global Navigation Satellite Systems (GNSS) chipsets, inertial sensors, barometer, lidar, radar and vision sensors. The two sensors used most often by drone autopilot controllers for absolute positioning are the GNSS chipsets and barometer. Although, for most outdoor operations, these sensors provide accurate and reliable position information, their accuracy, availability, and integrity deteriorate for indoor applications and in the presence of radio frequency interference (RFI), such as GNSS spoofing and jamming. Therefore, it is possible to derive network-based locations from Wi-Fi and cellular transmission. Although there have been many theoretical studies on network positioning, limited resources are available for the expected quantitative performance of these positioning methodologies. In this paper, the authors investigate both the horizontal and vertical accuracy of the Android network location engines under rural, suburban, and urban environments. The paper determines the horizontal location accuracy to be approximately 1637 m, 38 m, and 32 m in terms of 68% circular error probable (CEP) for rural, suburban, and urban environments, respectively, and the vertical accuracy to be 1.2 m and 4.6 m in terms of 68% CEP for suburban and urban environments, respectively. In addition, the availability and latency of the location engines are explored. Furthermore, the paper assesses the accuracy of the Android network location accuracy indicator for various drone operation environments. The assessed accuracies of the network locations provide a deeper insight into their potential for drone navigation.

Developing Drone AI to Avoid Risks Using Deep Learning

To improve the reliability and safety of autonomous drone navigation, we are conducting research and development with the aim of realizing "no-fall/safe-to-fall" drones equipped with deep learning AI technology. It consists of autonomous operation AI technology that detects people and vehicles and avoids flying over them, failure diagnosis AI technology that detects malfunctions and unstable behavior based on onboard drone sensors and telemetry information, and emergency landing AI technology that moves to a safe area and lands when a failure is diagnosed. Techniques for finding a safe area for a drone to land during an emergency landing and approaches to ensure object detection performance are described.

Learning in AI Processor

AI processor, which can run artificial intelligence algorithms, is a state-of-the-art accelerator,in essence, to perform special algorithm in various applications. In particular,these are four AI applications: VR/AR smartphone games, high-performance computing, Advanced Driver Assistance Systems and IoT. Deep learning using convolutional neural networks (CNNs) involves embedding intelligence into applications to perform tasks and has achieved unprecedented accuracy [1]. Usually, the powerful multi-core processors and the on-chip tensor processing accelerator unit are prominent hardware features of deep learning AI processor. After data is collected by sensors, tools such as image processing technique, voice recognition and autonomous drone navigation, are adopted to pre-process and analyze data. In recent years, plenty of technologies associating with deep learning Al processor including cognitive spectrum sensing, computer vision and semantic reasoning become a focus in current research.

Flying Free: A Research Overview of Deep Learning in Drone Navigation Autonomy

With the rise of Deep Learning approaches in computer vision applications, significant strides have been made towards vehicular autonomy. Research activity in autonomous drone navigation has increased rapidly in the past five years, and drones are moving fast towards the ultimate goal of near-complete autonomy. However, while much work in the area focuses on specific tasks in drone navigation, the contribution to the overall goal of autonomy is often not assessed, and a comprehensive overview is needed. In this work, a taxonomy of drone navigation autonomy is established by mapping the definitions of vehicular autonomy levels, as defined by the Society of Automotive Engineers, to specific drone tasks in order to create a clear definition of autonomy when applied to drones. A top–down examination of research work in the area is conducted, focusing on drone navigation tasks, in order to understand the extent of research activity in each area. Autonomy levels are cross-checked against the drone navigation tasks addressed in each work to provide a framework for understanding the trajectory of current research. This work serves as a guide to research in drone autonomy with a particular focus on Deep Learning-based solutions, indicating key works and areas of opportunity for development of this area in the future.

VGF-Net: Visual-Geometric fusion learning for simultaneous drone navigation and height mapping

The drone navigation requires the comprehensive understanding of both visual and geometric information in the 3D world. In this paper, we present a Visual-Geometric Fusion Network (VGF-Net), a deep network for the fusion analysis of visual/geometric data and the construction of 2.5D height maps for simultaneous drone navigation in novel environments. Given an initial rough height map and a sequence of RGB images, our VGF-Net extracts the visual information of the scene, along with a sparse set of 3D keypoints that capture the geometric relationship between objects in the scene. Driven by the data, VGF-Net adaptively fuses visual and geometric information, forming a unified Visual-Geometric Representation. This representation is fed to a new Directional Attention Model (DAM), which helps enhance the visual-geometric object relationship and propagates the informative data to dynamically refine the height map and the corresponding keypoints. An entire end-to-end information fusion and mapping system is formed, demonstrating remarkable robustness and high accuracy on the autonomous drone navigation across complex indoor and large-scale outdoor scenes.

A Simulation Framework for Developing Autonomous Drone Navigation Systems

Unmanned aerial vehicles are gaining popularity in an ever-increasing range of applications, mainly because they are able to navigate autonomously. In this work, we describe a simulation framework that can help engineering students who are starting out in the field of aerial robotics to acquire the necessary competences and skills for the development of autonomous drone navigation systems. In our framework, drone behavior is defined in a graphical way, by means of very intuitive state machines, whereas low-level control details have been abstracted. We show how the framework can be used to develop a navigation system proposal according to the rules of the “ESII Drone Challenge” student competition. We also show how the proposal can be evaluated in different test scenarios.