Autonomous Ground Vehicle Navigation Research Articles

Development of Local Path Planning Using Selective Model Predictive Control, Potential Fields, and Particle Swarm Optimization

This paper focuses on the real-time obstacle avoidance and safe navigation of autonomous ground vehicles (AGVs). It introduces the Selective MPC-PF-PSO algorithm, which includes model predictive control (MPC), Artificial Potential Fields (APFs), and particle swarm optimization (PSO). This approach involves defining multiple sets of coefficients for adaptability to the surrounding environment. The simulation results demonstrate that the algorithm is appropriate for generating obstacle avoidance paths. The algorithm was implemented on the ROS platform using NVIDIA’s Jetson Xavier, and driving experiments were conducted with a steer-type AGV. Through measurements of computation time and real obstacle avoidance experiments, it was shown to be practical in the real world.

Trajectory optimization with hybrid probabilistic roadmap approach to achieve time efficient navigation of unmanned vehicles in unstructured environment

PurposeThis paper aims to focus on solving the path optimization problem by modifying the probabilistic roadmap (PRM) technique as it suffers from the selection of the optimal number of nodes and deploy in free space for reliable trajectory planning.Design/methodology/approachTraditional PRM is modified by developing a decision-making strategy for the selection of optimal nodes w.r.t. the complexity of the environment and deploying the optimal number of nodes outside the closed segment. Subsequently, the generated trajectory is made smoother by implementing the modified Bezier curve technique, which selects an optimal number of control points near the sharp turns for the reliable convergence of the trajectory that reduces the sum of the robot’s turning angles.FindingsThe proposed technique is compared with state-of-the-art techniques that show the reduction of computational load by 12.46%, the number of sharp turns by 100%, the number of collisions by 100% and increase the velocity parameter by 19.91%.Originality/valueThe proposed adaptive technique provides a better solution for autonomous navigation of unmanned ground vehicles, transportation, warehouse applications, etc.

Waypoint Generation in Satellite Images Based on a CNN for Outdoor UGV Navigation

Moving on paths or trails present in natural environments makes autonomous navigation of unmanned ground vehicles (UGV) simpler and safer. In this sense, aerial photographs provide a lot of information of wide areas that can be employed to detect paths for UGV usage. This paper proposes the extraction of paths from a geo-referenced satellite image centered at the current UGV position. Its pixels are individually classified as being part of a path or not using a convolutional neural network (CNN) which has been trained using synthetic data. Then, successive distant waypoints inside the detected paths are generated to achieve a given goal. This processing has been successfully tested on the Andabata mobile robot, which follows the list of waypoints in a reactive way based on a three-dimensional (3D) light detection and ranging (LiDAR) sensor.

Where am I heading? A robust approach for orientation estimation of autonomous agricultural robots

Reliable knowledge of the vehicle heading plays a significant role in the autonomous navigation of agricultural Unmanned Ground Vehicles (UGVs), especially in the context of unstructured outdoor environments such as rural and forestry scenarios. However, achieving this information with an acceptable degree of confidence is a non-trivial task and still an open field of research. Expensive solutions are available on the market, but they often discourage most farmers due to the large investments needed for the startup. This paper introduces a novel algorithmic solution for reliable evaluation of the absolute vehicle heading, grounded on adaptive Kalman filtering with input evaluation via linear regression analysis. The proposed approach provides a functional and affordable solution to the heading estimation problem that can be used in real-world applications. The system is validated through an extensive experimental campaign using an all-terrain tracked rover operating in agricultural settings, showing good accuracy compared to other approaches, such as a dual GPS method found in the literature.

LiDAR SLAM with a Wheel Encoder in a Featureless Tunnel Environment

Simultaneous localization and mapping (SLAM) represents a crucial algorithm in the autonomous navigation of ground vehicles. Several studies were conducted to improve the SLAM algorithm using various sensors and robot platforms. However, only a few works have focused on applications inside low-illuminated featureless tunnel environments. In this work, we present an improved SLAM algorithm using wheel encoder data from an autonomous ground vehicle (AGV) to obtain robust performance in a featureless tunnel environment. The improved SLAM system uses FAST-LIO2 LiDAR SLAM as the baseline algorithm, and the additional wheel encoder sensor data are integrated into the baseline SLAM structure using the extended Kalman filter (EKF) algorithm. The EKF algorithm is used after the LiDAR odometry estimation and before the mapping process of FAST-LIO2. The prediction step uses the wheel encoder and inertial measurement unit (IMU) data, while the correction step uses the FAST-LIO2 LiDAR state estimation. We used an AGV to conduct experiments in flat and inclined terrain sections in a tunnel environment. The results showed that the mapping and the localization process in the SLAM algorithm was greatly improved in a featureless tunnel environment considering both inclined and flat terrains.

Energy efficient path planning for autonomous ground vehicles with ackermann steering

The autonomous ground vehicles have attracted a great deal of attention as viable solutions to a wide variety of military and civilian applications. However, the energy consumption plays a major role in the navigation of autonomous ground vehicles in challenging environments, especially if they are left to operate unattended under limited on-board power, such as planetary exploration, border patrol, etc. The autonomous ground vehicles are expected to perform more tasks more efficiently with limited power in these scenarios. Although plenty of research has developed an effective methodology for generating dynamically feasible and energy efficient trajectories for skid steering or differential steering vehicles, few studies on path planning for ackermann steering autonomous ground vehicles are available. In this study, an energy efficient path planning method with guarantee on completeness is proposed for autonomous ground vehicle with ackermann steering which is based on A∗ search algorithm. Firstly, the energy cost model is established for the autonomous ground vehicle using its kinematic constraints. Then, given the start and goal states, the energy-aware motion primitives are generated offline using the energy cost model to calculate the cost of each primary trajectory. Lastly, the energy efficient path planner is proposed and the analysis for completeness properties is given. The effectiveness of the proposed energy efficient path planner is verified by simulation over 150 randomly generated maps and real vehicle tests. The results show that a small increase in the distance of a path over the distance optimal path can result in a reduction of energy cost by nearly 26.9% in simulation and 21.09% in real test scenario for autonomous ground vehicles with ackermann steering.

AutoVRL: A High Fidelity Autonomous Ground Vehicle Simulator for Sim-to-Real Deep Reinforcement Learning

Deep Reinforcement Learning (DRL) enables cognitive Autonomous Ground Vehicle (AGV) navigation utilizing raw sensor data without a-priori maps or GPS, which is a necessity in hazardous, information poor environments such as regions where natural disasters occur, and extraterrestrial planets. The substantial training time required to learn an optimal DRL policy, which can be days or weeks for complex tasks, is a major hurdle to real-world implementation in AGV applications. Training entails repeated collisions with the surrounding environment over an extended time period, dependent on the complexity of the task, to reinforce positive exploratory, application specific behavior that is expensive, and time consuming in the real-world. Effectively bridging the simulation to real-world gap is a requisite for successful implementation of DRL in complex AGV applications, enabling learning of cost-effective policies. We present AutoVRL, an open-source high fidelity simulator built upon the Bullet physics engine utilizing OpenAI Gym and Stable Baselines3 in PyTorch to train AGV DRL agents for sim-to-real policy transfer. AutoVRL is equipped with sensor implementations of GPS, IMU, LiDAR and camera, actuators for AGV control, and realistic environments, with extensibility for new environments and AGV models. The simulator provides access to state-of-the-art DRL algorithms, utilizing a python interface for simple algorithm and environment customization, and simulation execution.

Occupancy Grid Generation With Dynamic Obstacle Segmentation in Stereo Images

The detection of dynamic and static obstacles is a key task for the navigation of autonomous ground vehicles. The article presents a new algorithm for generating an occupancy map of the surrounding space from noisy point clouds obtained from one or several stereo cameras. The camera images are segmented by the proposed deep neural network FCN-ResNet-M-OC, which combines the speed of the FCN-ResNet method and improves the quality of the model using the concept of object context representation. The paper investigates supervised approaches to network training on unbalanced samples with road scenes such as the weighted cross entropy and the Focal Loss. The occupancy map is built from point clouds with semantic labels, in which static environment and potentially dynamic obstacles are highlighted. Our solution is operational in real time and applicable on platforms with limited computing resources. The approach was tested on autonomous vehicle datasets: Semantic KITTI, KITTI-360, Mapillary Vistas and custom OpenTaganrog. The usage of semantically labeled point clouds increased the precision of obstacle detection by an average of 17%. The performance of the entire approach on various computing platforms with Jetson Xavier, RTX3070, GPUs NVidia Tesla V100 is respectively from 10 to 15 FPS for input image resolution <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$1920\times 1080$ </tex-math></inline-formula> pixels.

Fast Route Planner Considering Terrain Information.

Route planning considering terrain information is useful for the navigation of autonomous ground vehicles (AGV) on complicated terrain surfaces, such as mountains with rivers. For instance, an AGV in mountains cannot cross a river or a valley that is too steep. This article addresses a novel route-planning algorithm that is time-efficient in building a sub-optimal route considering terrain information. In order to construct a route from the start to the end point in a time-efficient manner, we simulate two virtual vehicles that deploy virtual nodes iteratively, such that the connected node network can be formed. The generated node network serves as a topological map for a real AGV, and we construct the shortest route from the start to the end point utilizing the network. The route is weighted considering the route length, the steepness of the route, and the traversibility of the route. Through MATLAB simulations, we demonstrate the effectiveness of the proposed route-planning algorithm by comparing it with RRT-star planners.

An autonomous navigation approach for unmanned vehicle in outdoor unstructured terrain with dynamic and negative obstacles

AbstractAt present, the study on autonomous unmanned ground vehicle navigation in an unstructured environment is still facing great challenges and is of great significance in scenarios where search and rescue robots, planetary exploration robots, and agricultural robots are needed. In this paper, we proposed an autonomous navigation method for unstructured environments based on terrain constraints. Efficient path search and trajectory optimization on octree map are proposed to generate trajectories, which can effectively avoid various obstacles in off-road environments, such as dynamic obstacles and negative obstacles, to reach the specified destination. We have conducted empirical experiments in both simulated and real environments, and the results show that our approach achieved superior performance in dynamic obstacle avoidance tasks and mapless navigation tasks compared to the traditional 2-dimensional or 2.5-dimensional navigation methods.

Learning-Based Methods of Perception and Navigation for Ground Vehicles in Unstructured Environments: A Review.

The problem of autonomous navigation of a ground vehicle in unstructured environments is both challenging and crucial for the deployment of this type of vehicle in real-world applications. Several well-established communities in robotics research deal with these scenarios such as search and rescue robotics, planetary exploration, and agricultural robotics. Perception plays a crucial role in this context, since it provides the necessary information to make the vehicle aware of its own status and its surrounding environment. We present a review on the recent contributions in the robotics literature adopting learning-based methods to solve the problem of environment perception and interpretation with the final aim of the autonomous context-aware navigation of ground vehicles in unstructured environments. To the best of our knowledge, this is the first work providing such a review in this context.

Decentralised indoor smart camera mapping and hierarchical navigation for autonomous ground vehicles

In this work, the authors propose a novel decentralised coordination scheme for autonomous ground vehicles to enable map building and path planning with a network of smart overhead cameras. Decentralised indoor smart camera mapping and hierarchical navigation supports the automatic generation of waypoint graphs for each camera in an environment and allows path planning through the environment across multiple camera fields of view, or subviews. The proposed solution utilises the growing neural gas algorithm to learn the topology of unoccupied working space in each subview for maintaining a dynamic waypoint graph on each camera. The authors’ pathing solution leverages a modified version of the A* algorithm to compute paths in a decentralised and hierarchical fashion. Waypoint generation was simulated and analysed on a generated environment to ensure it is both effective and efficient, while path planning was simulated on various randomised hierarchical graphs to effectively compare the proposed Decentralised-A* (D-A*) algorithm against standard greedy search. The proposed method efficiently handles the cases where other robot navigation methods are otherwise weak and ineffective, while still providing avenues for further optimisation of resource overhead for both the smart camera network as well as the robots themselves.

Supervised Learning of Natural-Terrain Traversability with Synthetic 3D Laser Scans

Autonomous navigation of ground vehicles on natural environments requires looking for traversable terrain continuously. This paper develops traversability classifiers for the three-dimensional (3D) point clouds acquired by the mobile robot Andabata on non-slippery solid ground. To this end, different supervised learning techniques from the Python library Scikit-learn are employed. Training and validation are performed with synthetic 3D laser scans that were labelled point by point automatically with the robotic simulator Gazebo. Good prediction results are obtained for most of the developed classifiers, which have also been tested successfully on real 3D laser scans acquired by Andabata in motion.

Coverage path planning for a flock of aerial vehicles to support autonomous rovers through traversability analysis

In rough terrains, such as landslides or volcanic eruptions, it is extremely complex to plan safe trajectories for an Unmanned Ground Vehicle (UGV), since both robot stability and path execution feasibility must be guaranteed. In this paper, we present a complete solution for the autonomous navigation of ground vehicles in the mentioned scenarios. The proposed solution integrates three different aspects. The first is the coverage path planning for the definition of UAV trajectories for aerial imagery acquisition. The collected images are used for the photogrammetric reconstruction of the considered area. The second aspect is the adoption of a flock of UAVs to implement the coverage in a parallel way. In fact, when non-coverable zones are present, decomposition of the whole area to survey is performed. A solution to assign the different regions among the flying vehicles composing the team is presented. The last aspect is the path planning of the ground vehicle by means of a traversability analysis performed on the terrain 3D model. The computed paths are optimal in terms of the difficulty of moving across the rough terrain. The results of each step within the overall approach are shown.

Secondary Radar Beacons for Local Ad-Hoc Autonomous Robot Localization Systems.

In this paper, we present a detailed analysis and implementation of secondary radar beacons designed for a local ad-hoc localization and landing system (LAOLa) to support the navigation of autonomous ground and aerial vehicles. We discuss a switched linear feedback network as a virtually coherent oscillator and show how to use it as a secondary radar transponder. Further, we present a signal model for the beat signal of the transponder response in an FMCW radar system, which is more detailed than in previously published papers. An actual transponder realization in the 24 ISM band is presented. Its RF performance was evaluated both in the laboratory and in the field. Finally, we put forward some ideas on how to overcome the range measurement inaccuracy inherent in this transponder concept.

Sliding-Window Temporal Attention Based Deep Learning System for Robust Sensor Modality Fusion for UGV Navigation

We propose a novel temporal attention based neural network architecture for robotics tasks that involve fusion of time series of sensor data, and evaluate the performance improvements in the context of autonomous navigation of unmanned ground vehicles (UGVs) in uncertain environments. The architecture generates feature vectors by fusing raw pixel and depth values collected by camera(s) and LiDAR(s), stores a history of the generated feature vectors, and incorporates the temporally attended history with current features to predict a steering command. The experimental studies show the robust performance in unknown and cluttered environments. Furthermore, the temporal attention is resilient to noise, bias, blur, and occlusions in the sensor signals. We trained the network on indoor corridor datasets (that will be publicly released) from our UGV. The datasets have LiDAR depth measurements, camera images, and human tele-operation commands.

Optimized RRT-A* Path Planning Method for Mobile Robots in Partially Known Environment

This paper presents optimized rapidly exploring random trees A* (ORRT-A*) method to improve the performance of RRT-A* method to compute safe and optimal path with low time complexity for autonomous mobile robots in partially known complex environments. ORRT-A* method combines morphological dilation, goal-biased RRT, A* and cubic spline algorithms. Goal-biased RRT is modified by introducing additional step-size to speed up the generation of the tree towards the goal after which A* is applied to obtain the shortest path. Morphological dilation technique is used to provide safety for the robots while cubic spline interpolation is used to smoothen the path for easy navigation. Results indicate that ORRT-A* method demonstrates improved path quality compared to goal-biased RRT and RRT-A* methods. ORRT-A* is therefore a promising method in achieving autonomous ground vehicle navigation in unknown environments

Research on Time-Correlated Errors Using Allan Variance in a Kalman Filter Applicable to Vector-Tracking-Based GNSS Software-Defined Receiver for Autonomous Ground Vehicle Navigation

The global navigation satellite system (GNSS) has been applied to many areas, e.g.,the autonomous ground vehicle, unmanned aerial vehicle (UAV), precision agriculture, smart city,and the GNSS-reflectometry (GNSS-R), being of considerable significance over the past few decades.Unfortunately, the GNSS signal performance has the high risk of being reduced by the environmentalinterference. The vector tracking (VT) technique is promising to enhance the robustness in highdynamics as well as improve the sensitivity against the weak environment of the GNSS receiver.However, the time-correlated error coupled in the receiver clock estimations in terms of the VT loopcan decrease the accuracy of the navigation solution. There are few works present dealing with thisissue. In this work, the Allan variance is accordingly exploited to specify a model which is expectedto account for this type of error based on the 1st-order Gauss-Markov (GM) process. Then, it is usedfor proposing an enhanced Kalman filter (KF) by which this error can be suppressed. Furthermore,the proposed system model makes use of the innovation sequence so that the process covariancematrix can be adaptively adjusted and updated. The field tests demonstrate the performance of theproposed adaptive vector-tracking time-correlated error suppressed Kalman filter (A-VTTCES-KF).When compared with the results produced by the ordinary adaptive KF algorithm in terms of the VTloop, the real-time kinematic (RTK) positioning and code-based differential global positioning system(DGPS) positioning accuracies have been improved by 14.17% and 9.73%, respectively. On the otherhand, the RTK positioning performance has been increased by maximum 21.40% when comparedwith the results obtained from the commercial low-cost U-Blox receiver.

Deep Learning-Based GNSS Network-Based Real-Time Kinematic Improvement for Autonomous Ground Vehicle Navigation

Much navigation over the last several decades has been aided by the global navigation satellite system (GNSS). In addition, with the advent of the multi-GNSS era, more and more satellites are available for navigation purposes. However, the navigation is generally carried out by point positioning based on the pseudoranges. The real-time kinematic (RTK) and the advanced technology, namely, the network RTK (NRTK), were introduced for better positioning and navigation. Further improved navigation was also investigated by combining other sensors such as the inertial measurement unit (IMU). On the other hand, a deep learning technique has been recently evolving in many fields, including automatic navigation of the vehicles. This is because deep learning combines various sensors without complicated analytical modeling of each individual sensor. In this study, we structured the multilayer recurrent neural networks (RNN) to improve the accuracy and the stability of the GNSS absolute solutions for the autonomous vehicle navigation. Specifically, the long short-term memory (LSTM) is an especially useful algorithm for time series data such as navigation with moderate speed of platforms. From an experiment conducted in a testing area, the LSTM algorithm developed the positioning accuracy by about 40% compared to GNSS-only navigation without any external bias information. Once the bias is taken care of, the accuracy will significantly be improved up to 8 times better than the GNSS absolute positioning results. The bias terms of the solution need to be estimated within the model by optimizing the layers as well as the nodes each layer, which should be done in further research.

A deep learning gated architecture for UGV navigation robust to sensor failures

In this paper, we introduce a novel methodology for fusing sensors and improving robustness to sensor failures in end-to-end learning based autonomous navigation of ground vehicles in unknown environments. We propose the first learning based camera–LiDAR fusion methodology for autonomous in-door navigation. Specifically, we develop a multimodal end-to-end learning system, which maps raw depths and pixels from LiDAR and camera, respectively, to the steering commands. A novel gating based dropout regularization technique is introduced which effectively performs multimodal sensor fusion and reliably predicts steering commands even in the presence of various sensor failures. The robustness of our network architecture is demonstrated by experimentally evaluating its ability to autonomously navigate in the indoor corridor environment. Specifically, we show through various empirical results that our framework is robust to sensor failures, partial image occlusions, modifications of the camera image intensity, and the presence of noise in the camera or LiDAR range images. Furthermore, we show that some aspects of obstacle avoidance are implicitly learned (while not being specifically trained for it); these learned navigation capabilities are shown in ground vehicle navigation around static and dynamic obstacles.