2023· International Journal of Intelligent Automation & Robotics Engineering· 0 citations
TL;DR
This work forms a prior method based on Deep Deterministic Policy Gradient controller and a adaptive unscented Kalman filter which continuously providing constantly estimating robot states and improving the motion primitives in hazardous operating conditions to validate that end-to-end AI navigation architectures deliver the robust performance needed for next-gen automated industrial monitoring and non-destructive evaluation at scale.
Abstract
We focus on autonomous inspection robots operating in complex, dynamic, and GNSS-denied industrial environments dealing with critical problems related to real-time trajectory optimization, feature tracking and precise spatial localization. Remember that traditional control algorithms tend to not adapt well to difficult visual occlusions, non-Gaussian sensor noise, or unforeseen structural impediments. We propose a unified AI pipeline that fuses deep reinforcement learning with sensor data—by combining Light Detection and Ranging (LiDAR), Visual-Inertial Odometry (VIO), and thermal images—to discover flexible navigation strategies for autonomous inspection ground vehicles. In this work, we form a prior method based on Deep Deterministic Policy Gradient controller and a adaptive unscented Kalman filter which continuously providing constantly estimating robot states and improving the motion primitives in hazardous operating conditions. A series of experimental evaluations conducted on both simulated industrial plants and a physical mock-up facility show that the AI-based method achieves up to 51% relative reduction in localization error compared to traditional Simultaneous Localization and Mapping methods. The results show that the path deviation decrease by 38.15%, collision avoidance timeliness is significantly improved, and the accuracy of anomalies detection can reach more than 98%. These results validate that end-to-end AI navigation architectures deliver the robust performance needed for next-gen automated industrial monitoring and non-destructive evaluation at scale.
Simulation and physical experiments confirmed collision-free navigation and successful quick response (QR)-code-based goods inspection, demonstrating the feasibility of the proposed framework for small, structured indoor environments.
T. Q. Le, T. Luu· IAES International Journal o...· 0 citations
An intelligent vision-based autonomous robotic framework that integrates deep learning-based object detection with hybrid adaptive navigation for dynamic environments is proposed in this research. The proposed system addresses the challenges of real-time perception and robust navigation in unstructured settings by combining a convolutional neural network (CNN) for object detection with a hybrid control mechanism for motion planning. The CNN, implemented using a state-of-the-art architecture such as YOLO, processes visual input to identify obstacles and target objects, providing critical environmental awareness. Moreover, the hybrid navigation strategy merges reactive obstacle avoidance, achieved through algorithms like the Vector Field Histogram (VFH), with adaptive path planning using Rapidly-exploring Random Trees (RRT) to ensure both immediate collision avoidance and long-term goal convergence. The integration of these components enables the robotic system to dynamically adjust its navigation policy in response to environmental changes, thereby improving robustness and adaptability. The novelty of our approach lies in the seamless fusion of vision-based perception and adaptive control, which enhances the system’s capability to operate in complex, dynamic scenarios. Experimental validation demonstrates the effectiveness of the framework in real-world applications, highlighting its potential for deployment in autonomous vehicles, service robotics, and industrial automation. The proposed method offers a scalable and efficient solution for autonomous systems requiring high levels of situational awareness and adaptive decision-making.
K. A.· International Journal on Rob...· 0 citations
Unstructured robotic navigation is a critical research area due to its applications in disaster response, space exploration, agriculture, and military operations. Unlike structured environments, unstructured settings are unpredictable, dynamic, and lack complete sensory information, making navigation highly complex. Before 2018, research focused on classical and early intelligent methods such as probabilistic robotics, heuristic path planning, and initial machine learning integration. Key navigation tasks—localization, mapping, path planning, and motion control—were addressed using techniques like Bayesian filtering, Kalman filters, particle filters, and occupancy grid mapping to handle uncertainty. Algorithms such as A*, D*, and Rapidly-exploring Random Trees (RRT) were widely used for path planning, often enhanced with heuristics and real-time replanning for dynamic environments. Sensor fusion combining LiDAR, sonar, and vision improved environmental perception, while early AI approaches like neural networks and fuzzy logic enabled adaptive decision-making. Reinforcement learning also showed potential, though it was limited by computational constraints at the time.Despite significant progress, pre-2018 systems faced challenges such as limited computational power, poor generalization, and reliance on handcrafted features. Overall, these foundational methods played a vital role in advancing autonomous navigation, though achieving full autonomy in complex environments remains an ongoing challenge.
Lakshmi Narayanan· International Journal of Int...· 0 citations
To overcome the severe perceptual sparsity of pipeline interiors, this paper presents an active Visual-Inertial Odometry (VIO) framework that generates its own visual landmarks on the fly. By utilizing a pulse of compressed air through a modified airbrush, the robotic platform deposits non-uniform, non-permanent fluorescent markings onto the pipe walls, creating reliable visual features for a dual-camera front-end. Alongside this active perception strategy, the system employs a multi-IMU sensor suite optimized specifically for the cylindrical manifold of pipe environments. We demonstrate via Fisher Information analysis that a closed-form, optimally placed IMU configuration significantly enhances the observability of the robot’s motion, proving that geometric arrangement dominates over mere sensor count. Validated across diverse pipe geometries and surface textures in both simulation and real-world experiments, the proposed VIO approach consistently eliminates tracking failures, achieves lower trajectory RMSE than visual-only and sub-optimal inertial baselines, and keeps the residual errors strictly contained within the pipe corridor.
Aristeidis Geladaris, Athanasios S. Mastrogeorgiou, Odysseas Simatos et al.· IEEE Access· 0 citations
Autonomous mobile robots operating in dynamically changing, unstructured environments require high-precision, drift-free localization capabilities to achieve robust operational safety and navigational efficacy. While visual Simultaneous Localization and Mapping (vSLAM) and Inertial Navigation Systems (INS) serve as foundational technologies in intelligent automation, standalone implementations encounter significant vulnerabilities, specifically optical occlusion and cumulative dead-reckoning drift. This paper presents a comprehensive study on an intelligent, optimization-based, tightly-coupled vision-inertial sensor fusion framework designed for robust localization in challenging environments. The proposed system integrates high-frequency inertial measurements from an Inertial Measurement Unit (IMU) with high-fidelity visual landmarks extracted from a monocular camera, utilizing an artificial intelligence-driven adaptive Extended Kalman Filter (EKF) state estimation matrix to dynamically adjust measurement noise weights. By analyzing the structural characteristics of feature tracking alongside high-frequency acceleration profiles, the intelligent layer dampens sensor anomalies caused by aggressive motion or lightning fluctuations. Experimental validations conducted using the EuRoC MAV public benchmark dataset indicate that the proposed intelligent fusion architecture provides superior performance across dynamic trajectories, reducing the Absolute Trajectory Error (ATE) by up to 34% compared to classical loosely-coupled filtering methods while maintaining sub-centimeter positional drift thresholds.
J. Arsac· International Journal of Int...· 0 citations
Complex and dynamic environment autonomy, the ability to navigate complex and dynamic environments autonomously in real-time and with limited computational power, will persist as a basic challenge to mobile robotic systems. As a solution to this problem, this paper presents an alternative minimum-cost deep learning format using optimization, which combines a Modified Rime Optimization Algorithm (M-RIME) and a lightweight SqueezeNet-based navigation network. The given strategy is supposed to increase the depth of perception, precision of motion control, and computational efficiency at the same time. To begin with, an end-to-end navigation model that uses SqueezeNet is created to concurrently map raw red–green–blue observations to continuous motion commands to facilitate the perception–action interface. Subsequently, the M-RIME algorithm is improved with adaptive exploration–exploitation balancing and an adaptive mutation operator that greatly enhance convergence speed, global exploration, and the stability of solutions. The presented optimizer is used to improve the performance of the SqueezeNet model by automatically optimizing its hyperparameters within an offline optimization system, which improves navigation and retains its real-time functionality. Multi-frame and multi-measurement simulation experiments are performed in complex indoor navigation environments, where success rate, collision rate, navigation time, path length, and frames per second are adopted as quantitative and qualitative metrics to evaluate the suggested framework. The findings show that the optimized SqueezeNet is not only much better than the baseline model and other state-of-the-art optimization frameworks in terms of success rates of navigation, smoother trajectories, lower rates of collisions, and high-quality real-time optimization but also significantly better. In addition, convergence analysis establishes the effectiveness and precision of the suggested M-RIME algorithm in a high-dimensional optimization setting. On the whole, the presented framework will offer a good and scalable choice for autonomous navigation in real-time and can serve as an excellent basis for future developments aimed at temporal modeling, practical applications of autonomous driving, and multi-sensor fusion-driven robotic perception systems.
Jiangjun Ruan, Mei Wang, Lu Peng et al.· AIP Advances· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
