Autonomous Robot Research Articles

Night Vision Security Patrolling Robot with Sound Sensing

This project presents a Raspberry Pi-based automated system designed to integrate night vision and audio sensing capabilities for versatile applications such as surveillance and robotics. The system employs a night vision camera for visual data acquisition and a microphone for sound sensing, enabling real-time monitoring and intelligent responses to environmental stimuli. The Raspberry Pi serves as the central processing unit, seamlessly interfacing with the camera, microphone, and motor driver. The motor driver controls DC motors with specifications of 12V and 30 RPM, facilitating precise mechanical movements required for system operation. Powered by an efficient power supply, the system ensures consistent functionality even in resource-constrained environments. The modular design allows adaptability for a wide range of applications, including home security, autonomous robotics, and remote surveillance in low-light conditions. With the integration of a motor driver, the system enables automated movements, making it suitable for mobile applications such as patrolling robots or automated guided vehicles (AGVs). This project emphasizes the practicality and scalability of utilizing Raspberry Pi for cost-effective and efficient solutions in fields requiring simultaneous video and audio processing. Future enhancements could include adding artificial intelligence (AI) for object and sound recognition, wireless connectivity for remote monitoring, and advanced motor controls for more complex operations. The proposed system is a promising step toward accessible and adaptable technologies that combine vision, sound, and mobility in a compact and easy-to-deploy platform. This work contributes to the growing field of IoT-enabled smart systems and robotics by demonstrating an efficient multi-functional solution.

R2SCAT-LPR: Rotation-Robust Network with Self- and Cross-Attention Transformers for LiDAR-Based Place Recognition

LiDAR-based place recognition (LPR) is crucial for the navigation and localization of autonomous vehicles and mobile robots in large-scale outdoor environments and plays a critical role in loop closure detection for simultaneous localization and mapping (SLAM). Existing LPR methods, which utilize 2D bird’s-eye view (BEV) projections of 3D point clouds, achieve competitive performance in efficiency and recognition accuracy. However, these methods often struggle with capturing global contextual information and maintaining robustness to viewpoint variations. To address these challenges, we propose R2SCAT-LPR, a novel, transformer-based model that leverages self-attention and cross-attention mechanisms to extract rotation-robust place feature descriptors from BEV images. R2SCAT-LPR consists of three core modules: (1) R2MPFE, which employs weight-shared cascaded multi-head self-attention (MHSA) to extract multi-level spatial contextual patch features from both the original BEV image and its randomly rotated counterpart; (2) DSCA, which integrates dual-branch self-attention and multi-head cross-attention (MHCA) to capture intrinsic correspondences between multi-level patch features before and after rotation, enhancing the extraction of rotation-robust local features; and (3) a combined NetVLAD module, which aggregates patch features from both the original feature space and the rotated interaction space into a compact and viewpoint-robust global descriptor. Extensive experiments conducted on the KITTI and NCLT datasets validate the effectiveness of the proposed model, demonstrating its robustness to rotation variations and its generalization ability across diverse scenes and LiDAR sensors types. Furthermore, we evaluate the generalization performance and computational efficiency of R2SCAT-LPR on our self-constructed OffRoad-LPR dataset for off-road autonomous driving, verifying its deployability on resource-constrained platforms.

Design and Experimental Validation of an Effective Controller for Autonomous Differential Wheeled Robots

This research focuses on the control of differential wheeled robots and aims to develop a new, practical controller inspired by the well-known PID controller. The primary challenge with traditional PID controllers lies in the difficulty of accurately optimizing the three key parameters (coefficients of Proportional, Integral and Derivative) through experimental trial and error. While numerical simulations offer solutions for parameter optimization, real-world factors such as friction losses, inconsistent current and voltage supply to the DC motors, and wheel slippage often prevent these optimized parameters from achieving the same effectiveness in physical systems. This study proposes a novel controller designed to address these challenges, simplifying the parameter tuning process to a single, easily adjustable variable. The new controller, while inspired by the PID approach, offers an effective alternative, overcoming the limitations posed by real-world conditions. It was implemented and tested on a differential wheeled robot specifically designed for this research. Experimental testing across various tasks demonstrated that the new controller provides stable and reliable control, making it a valuable alternative to traditional PID-based approaches.

Edge-Cloud Orchestration of Assertion-Based Monitors for Robotic Applications

The runtime verification of multi-domain software applications implementing the behaviors of modern robots is a challenging task. On the one hand, assertion-based verification (ABV) has shown great potential to check the correctness of complex systems at runtime. On the other hand, the computational overhead introduced by runtime ABV can be substantial, variable and non-deterministic. As a consequence, applying accurate ABV at runtime to autonomous robots, which are often characterized by resource-constrained computing architectures, can lead to severe slowdowns of the software execution and failures of temporal constraints, thus compromising the overall system’s correctness. We address this challenge by proposing a platform for runtime ABV that implements monitor synthesis from signal temporal logic assertions and dynamic monitor migration across edge devices and the cloud. The synthesized monitors are wrapped into ROS-compliant nodes and connected to the system under verification. The overall ABV framework and the related migration mechanism are then containerized with Docker for both edge and cloud computing. To evaluate the proposed platform, we present the results obtained with a set of synthetic benchmarks and with an industrial case study, which implements the mission of a Robotnik RB-Kairos mobile robot in a smart manufacturing production line. Note to Practitioners . This paper was motivated by the need for accurate and runtime verification of robotic systems software. Verification and validation of intelligent systems are often incomplete, as they cannot anticipate all potential scenarios, including errors or unexpected events. On top of this, assertion-based verification can also be resource-intensive; therefore, careful use of resources is required to avoid overloading the robot’s computational resources with the monitors. To achieve this, we used signal temporal logic, a widely accepted solution to monitor robotic and distributed applications. The main contribution of this work is a framework that can automatically synthesize the monitors that interface with the Robot Operating System (ROS) and also the capability of optimizing the end-to-end latency of verification at runtime by exploiting a distributed computing architecture (i.e., edge-cloud). In future work, we will address not only the minimization of end-to-end latency but also the timing upper bound of monitors to achieve runtime deterministic verification.

Experimental study on the effects of ambient light on the precision of short-range near-infrared LiDAR systems

This work investigates the influence of ambient light noise on the performance of an 850 nm single-point light detection and ranging (LiDAR) system. It specifically focuses on different targets with varying reflectivity and range. A commercial LiDAR module with six distinct specimens was used to collect and analyze data over a measurement range of up to 4 m. The analysis of experimental data reveals that luminous intensity and range critically affect LiDAR accuracy. Reflectivity consistently affects LiDAR measurements, regardless of their magnitude, but it varies with the type of material. The relative error in measurement, the reflectivity of material, and the light noise of definite luminous intensity exhibit near-normal distributions with asymmetrical tails. Differences between the distributions of relative error, reflectivity, and luminous intensity highlight the complex interactions influencing LiDAR performance. These findings aid in the design and optimization of LiDAR systems, ensuring accurate environmental sensing under different ambient light conditions. Further, this data may be used to develop transferable models. Also, the findings of this study improve the performance of autonomous mobile robots for obstacle detection and collision avoidance in warehouses and in 3D printing and machining, where metallic surface reflectivity impacts LiDAR functioning.

Perception Systems for Autonomous Mobile Robots: Selecting and Mitigating Limits

A well-designed perception system is crucial for the proper operation of an autonomous mobile robot however it is not trivial to engineer. For a comprehensive system not only the sensors’ features must be considered but also the way their data are utilised for robotic operation. In this paper, we present a set of tightly coupled strategies that allow the creation of an end-to-end perception system utilising 3D representation in the form of point clouds. Our proposal is generic enough to be applied to various mobile robots as it is aware of the context of robotic operation and accounts for hardware constraints. As the first strategy, we introduce a formalised sensors selection process formulated as a multi-objective optimization problem with metrics relaxation which allows to choose an optimal set of sensors within budgetary limitations. Secondly, we validate various data filtration strategies and their combination in the context of the navigation system to find a trade-off between accuracy and computational effort. Finally, to mitigate the suboptimal field of view of combined sensors and augment the perception system we introduce a new concept for the occupancy grid layer which utilises motion information in the occupancy calculation. For these strategies, we conduct experimental verification and apply the results in an autonomous cleaning robot.

A comprehensive overview of Construction 4.0 technologies and their implementation in the construction industry

Purpose The purpose of this study is to examine Construction 4.0 trends, identify potential areas of growth, and explore the use of Industry 4.0 enabling technologies in the construction sector. Design/methodology/approach Research papers from referred international journals are reviewed to identify the use of Industry 4.0 enabling technologies in the construction sector. Then, Visualization of Similarities viewer software is used to analyze the keyword co-occurrence network, overlay and density visualization. Finally, the roadmap for the adoption of the Industry 4.0 concept in the construction sector is proposed. Findings Thirteen technologies of the Industry 4.0 concept have an impact on the construction sector. However, Digital Twins, the internet of things, Smart Sensors and 3D printing have the potential to support the Construction 4.0 concept, whereas applications of machine learning, autonomous robotics and augmented/virtual reality still offer plenty of opportunities for future applied research. Practical implications This research paper will help decision-makers in the construction sector to adopt the Industry 4.0 enabling technologies. Also, it will serve as a road map for future research in Construction 4.0. Originality/value The literature studied for the content analysis includes the years 2013 through 2023, which aids in the development of plans by policymakers for the adoption of the digital construction sector.

FPGA implementation of double-head SalsaNext: a CNN-based model for LiDAR point cloud segmentation

This study details the adaptation and deployment of a customized SalsaNext model for semantic segmentation of LiDAR point clouds on edge devices, benchmarked using the SemanticKITTI and Waymo Open datasets. We introduce an innovative multi-dataset training framework designed specifically for range image-based segmentation models. Central to this approach is our double-head SalsaNext model, which features two output heads to facilitate simultaneous training and inference on the Waymo and SemanticKITTI datasets. Following training, the model is streamlined by removing the head dedicated to Waymo, resulting in a compact, single-headed version optimized for SemanticKITTI. This simplified model is then quantized to employ fixed-point arithmetic, significantly enhancing computational efficiency and enabling real-time operation on the Xilinx Kria KV260 board. The quantization process markedly reduces resource consumption while preserving competitive accuracy. Our deployment on this low-power, FPGA-based platform underscores the potential of energy-efficient systems for advanced 3D semantic segmentation, with promising applications in autonomous systems and robotics. Experimental results validate the effectiveness of our training schema and the success of the optimized implementation of the double-head model on resource-constrained hardware.

Development and Evaluation of a Multiaxial Modular Ground Robot for Estimating Soybean Phenotypic Traits Using an RGB-Depth Sensor

Achieving global sustainable agriculture requires farmers worldwide to adopt smart agricultural technologies, such as autonomous ground robots. However, most ground robots are either task- or crop-specific and expensive for small-scale farmers and smallholders. Therefore, there is a need for cost-effective robotic platforms that are modular by design and can be easily adapted to varying tasks and crops. This paper describes the hardware design of a unique, low-cost multiaxial modular agricultural robot (ModagRobot), and its field evaluation for soybean phenotyping. The ModagRobot’s chassis was designed without any welded components, making it easy to adjust trackwidth, height, ground clearance, and length. For this experiment, the ModagRobot was equipped with an RGB-Depth (RGB-D) sensor and adapted to safely navigate over soybean rows to collect RGB-D images for estimating soybean phenotypic traits. RGB images were processed using the Excess Green Index to estimate the percent canopy ground coverage area. 3D point clouds generated from RGB-D images were used to estimate canopy height (CH) and the 3D Profile Index of sample plots using linear regression. Aboveground biomass (AGB) was estimated using extracted phenotypic traits. Results showed an R2, RMSE, and RRMSE of 0.786, 0.0181 m, and 2.47%, respectively, between estimated CH and measured CH. AGB estimated using all extracted traits showed an R2, RMSE, and RRMSE of 0.59, 0.0742 kg/m2, and 8.05%, respectively, compared to the measured AGB. The results demonstrate the effectiveness of the ModagRobot for in-row crop phenotyping.

Multi-scale parallel gated local feature transformer

Visual Simultaneous Localization and Mapping (VSLAM) is a crucial technology for autonomous mobile vision robots. However, existing methods often suffer from low localization accuracy and poor robustness in scenarios with significant scale variations and low-texture environments, primarily due to insufficient feature extraction and reduced matching precision. To address these challenges, this paper proposes an improved multi-scale local feature matching algorithm based on LoFTR, named MSpGLoFTR. First, we introduce a Multi-Scale Local Attention Module (MSLAM), which achieves feature fusion and resolution alignment through multi-scale window partitioning and a shared multi-layer perceptron (MLP). Second, a Multi-Scale Parallel Attention Module is designed to capture features across various scales, enhancing the model’s adaptability to large-scale features and highly similar pixel regions. Finally, a Gated Convolutional Network (GCN) mechanism is incorporated to dynamically adjust weights, emphasizing key features while suppressing background noise, thereby further improving matching precision and robustness. Experimental results demonstrate that MSpGLoFTR outperforms LoFTR in terms of matching precision, relative pose estimation performance, and adaptability to complex scenarios. Notably, it excels in environments with significant illumination changes, scale variations, and viewpoint shifts. This makes MSpGLoFTR an efficient and robust feature matching solution for complex vision tasks.

Autonomous Search and Rescue Robot

- In disaster-stricken environments, traditional search and rescue methods often encounter significant challenges due to adverse conditions such as smoke, fog, darkness, or complex terrains. Advancements in autonomous robotics and sensor technologies provide effective solutions to enhance the efficiency and reliability of rescue operations. This work focuses on the development of an autonomous search and rescue robot equipped with LiDAR based mapping and thermal imaging technology. The LiDAR system enables real-time 2D mapping of the environment, ensuring precise navigation and obstacle avoidance in dynamic and unstructured terrains. At the same time, the thermal imaging system detects heat signatures, allowing the identification and localization of humans or animals that may be obscured by environmental hazards. The proposed robotic system is designed to operate in disaster scenarios where human intervention is difficult or hazardous. By utilizing LiDAR for navigation and thermal imaging for survivor detection, the robot can autonomously move through challenging terrains while effectively identifying potential survivors. The integration of advanced sensing technologies with intelligent navigation algorithms enhances adaptability and reliability in critical situations. This approach aims to improve the speed and accuracy of locating victims, ultimately contributing to more effective and life-saving search and rescue missions. Key Words: Autonomous, ROS Neotic, LIDAR, Teleoperation, Ubuntu Mate, Ubuntu Server, Navigation

Development and evaluation of a soft end effector for kiwifruit harvesting

ABSTRACT With the continuous expansion of the kiwifruit industry and the decrease in workforce due to the COVID-19 pandemic, the development of autonomous harvesting robots is of significant importance. Current harvesting robots' end-effectors face challenges such as lack of flexibility and the potential to damage fruits. This study proposes a soft picking end-effector that effectively addresses these issues. The soft fingers achieve clamping of the kiwifruit by bending through inflation. Comprehensive assessments of its compliance and fingertip force have been conducted through finite element analysis and experiments. In a free state, applying 90 KPa of air pressure can enable the fingers to achieve bending of over 110° Under a pulling picking method, applying an air pressure of 115 KPa can provide a fruit gripping force of 3.5N, sufficient to separate the fruit from the stem. Experiments in field conditions have shown that the average picking time per kiwifruit is 6.7s, with fruit detachment rate, harvesting rate, and damage rate at 90.90%, 86.36% and 5.00%, respectively, and the fruit gripping damage rate is 0%. The kiwifruit soft-bodied end-effector proposed in this study provides a reference for the entire non-destructive harvesting process of kiwifruit.

Editorial for the special issue on Advanced Technology in autonomous robots and swarm intelligence

Editorial for the special issue on Advanced Technology in autonomous robots and swarm intelligence

Online health-aware energy management strategy of a fuel cell hybrid autonomous mobile robot under startup–shutdown condition

Online health-aware energy management strategy of a fuel cell hybrid autonomous mobile robot under startup–shutdown condition

Development of an autonomous chess robot system using computer vision and deep learning

Development of an autonomous chess robot system using computer vision and deep learning

AI-based approaches for improving autonomous mobile robot localization in indoor environments: A comprehensive review

AI-based approaches for improving autonomous mobile robot localization in indoor environments: A comprehensive review

RL-based Variable Horizon Model Predictive Control of Multi-Robot Systems in Dynamic Environments

Abstract Multi-robot systems (MRS) consist of multiple autonomous robots that collaborate to perform tasks more efficiently than single-robot systems. These systems enhance flexibility, enabling applications in areas such as environmental monitoring, search and rescue, and agricultural automation while addressing challenges related to coordination, communication, and task assignment. Model Predictive Control (MPC) stands out as a promising controller for multi-robot control due to its preview capability and effective constraint handling. However, MPC's performance heavily relies on the chosen length of the prediction horizon. Extending the prediction horizon significantly raises computation costs, making its tuning time-consuming and task-specific. To address this challenge, we introduce a framework utilizing a Collective Reinforcement Learning strategy to generate the prediction horizon dynamically based on the states of the robots. We propose that the prediction horizon of any robot in MRS depends on the states of all the robots. Additionally, we propose a Versatile On-demand Collision Avoidance (VODCA) strategy to enable on-the-fly collision avoidance for multiple robots operating under varying prediction horizons. This approach establishes a better trade-off between performance and computation costs, allowing for adaptable prediction horizons for each robot at every time step. Numerical studies are performed to investigate the scalability of the proposed framework, the stiffness of the learned RL policy, and the comparison with the fixed horizon and existing variable horizon MPC methods. The framework is also implemented on multiple TurtleBot3 Waffle Pi for various multi-robot tasks.

A Review on Autonomous Robot for Efficient Multitasking Operations

A Review on Autonomous Robot for Efficient Multitasking Operations

Autonomous Soft Robots: Self-regulation, Self-sustained, and Recovery Strategies

Autonomous Soft Robots: Self-regulation, Self-sustained, and Recovery Strategies

Design of a Dual-Function Autonomous Disinfection Robot with Safety Filter-Based Motion Control

Design of a Dual-Function Autonomous Disinfection Robot with Safety Filter-Based Motion Control