International Journal of Robotics and Automation Technology

Computer-Vision Unmanned Aerial Vehicle Detection System Using YOLOv8 Architectures

Aleksandar Petrovic — 2024-05-22

Abstract: This work aims to test the performance of the you only look once version 8 (YOLOv8) model for the problem of drone detection. Drones are very slightly regulated and standards need to be established. With a robust system for detecting drones the possibilities for regulating their usage are becoming realistic. Five different sizes of the model were tested to determine the best architecture size for this problem. The results indicate high performance across all models and that each model is to be used for a specific case. Smaller models are suited for lightweight system approaches where some false identification is tolerable, while the largest models are to be used with stationary systems that require the best precision.

Concept Design of Reproductive Worm Robot: A Natural Inheritance of C.elegans Biological Worm

Debanik Roy — 2024-08-21

Abstract: Although research in the field of worm robotics has taken some stride in recent past, the connotation of such designs was missing the feel of biology in true sense hitherto. Design and firmware of bio-inspired robots have been attempted by several research groups globally but none of those seriously intrude different physiologoical systems of the said biological specimen. Our attempt on technology-driven ideation of the reproductive system of a celebrated biological worm, namely, Caenorhabditis elegans (C.elegans) has culminated into concept-designs of Reproductive Worm Robot. Incidentally, C.elegans is an interesting biological entity that evokes imagination and assertion to create miniature robotic systems, especially by mimicing its reproductive system. In this paper, we have reported the technological concept designs as well as part-hardware of the working prototype of representative Reproductive Worm Robots by naturally inheriting the reproductive mechanism of the biologocal C.elegans worm (notwithstanding the size effect).

Computer Vision Based Areal Photographic Rocket Detection using YOLOv8 Models

Luka Jovanovic — 2024-09-09

Advances in aerospace engineering and aerodynamics have pioneered space exploration and helped support telecommunication infrastructure. But these same developments have also aided in the creation of weapons of devastating impact. This necessitates the development of ways for detecting and tracking rockets. While several methods, mostly based on Doppler radar exist, the need for active radio emissions limits the applicability of these systems. A passive system has several advantages over traditional techniques, however their potential is largely unexplored. This work seeks to tackle this research gap by exploring the potential of emerging computer vision townies applied to rocket detection and tracking. The advantages of such a system are the relatively low cost as well as passive nature making observation stations harder to detect and easier to deploy. This work explores the potential of pre-trained, lightweight YOLOv8 architectures for rocket detection in real-world situations. A publicly available dataset is utilized and a comparative analysis is carried out between nano and small models. Both models demonstrate favorable outcomes with an accuracy of 0.90 for rocket body detection and 0.93 for engine flame detection. Nevertheless, rocket detection into space is still difficult, with a precision of 0.64 for this class. This paper indicates areas for additional refinement and demonstrates the potential of computer vision technology in passive rocket detection.

Self-Organized Multi-Robot Path Planning and Chain Distribution Based on Improved DWA and A* Fusion in Unknown Space

Bowen Ding — 2024-10-01

With the rise of autonomous driving in recent years, path planning has gained widespread attention. Traditional path planning needs to be based on a large amount of known information, which is not available for confined environments. Taking the complex indoor space where GPS cannot be used as the research background, the article designs a self-organised motion scheme for multi-intelligent body trolleys that includes exploration and path planning. By improving the DWA and A* algorithms, the multi-robot self-organisation achieves reasonable path planning, and the fusion of the two algorithms solves the contradictory problems of global planning being unable to avoid dynamic obstacles and local planning possibly falling into local optimum. After that, the pilot-following algorithm is added to guide the multi-intelligence body to operate in formation. By studying the constraints of hardware such as LiDAR and machine trolleys, the chain distribution of multiple intelligences is proposed to solve the problem of information loss caused by the discontinuous monitoring field of view. Eventually, when the carts are all in position, the whole area is covered and monitored using sensor fusion with multiple viewpoints. The feasibility of the explored scheme is verified by simulation experiments, and the feasibility and robustness of multi-sensor fusion is verified by specific hardware.

Lethal Autonomous Weapon Systems: Ethical Dilemmas and Legal Compliance in the Era of Military Disruptive Technologies

Marco Marsili — 2024-10-06

Lethal Autonomous Weapon Systems (LAWS) have emerged as one of the most significant advancements in military technology, leveraging artificial intelligence (AI) and machine learning to execute missions without direct human control. As these systems become central to modern warfare, they raise critical questions about their compliance with International Humanitarian Law (IHL) and International Human Rights Law (IHRL). This paper delves into the legal and ethical debates surrounding LAWS with particular attention to the discussions within the Group of Governmental Experts on LAWS (GGE on LAWS). We analyze whether these technologies can adhere to fundamental human rights while maintaining their operational efficacy. Through the application of the Autonomy Spectrum Framework to real-world scenarios, the study highlights both the strategic advantages of LAWS and the risks of dehumanizing warfare. The need for robust legal frameworks to ensure accountability and human oversight remains paramount.

Investigation and Fabrication of Brilliant Green Dye Material-Based Organic Heterojunction: Dielectric Response and Impedance Spectroscopy

M. Benhaliliba — 2024-10-15

Brilliant green (BG) dye material-based heterojunction is fabricated by a spin coating route and characterized by impedance analyzer. Here, we study the impedance spectroscopy of such organic material based thin film onto silicon substrate. The capacitance and conductance versus voltage (C-V), (G-V) characteristics are plotted. So, the dielectric response of BG based heterojunction is studied via the plotting of the dielectric constant, modulus components, complex impedance and Nyquist diagram. Real and imaginary parts of electrical conductivity are also plotted at several frequencies. Real and imaginary parts of electric modulus M’-V and M”-V are investigated for all experimental frequencies. The Z’-V characteristics of our device-based BG organic heterojunction exhibit a peak of 1383 W. It is confirmed that an anomalous peak of Z’ is recorded within 0.5-1 V range. The drastic decay in Z”-V plot occurs for lower 100 and 200 kHz frequency range and Z” values become constant beyond 1V for all frequencies. Ac and Dc conductivity curves demonstrate a growth with frequency and angle phase approaches to 90º within reverse voltage. This result reveals a capacitive conduct of our device.

Consumer Feedback Sentiment Classification Improved Via Modified Metaheuristic Optimization Natural Language Processing

Stanislava Kozakijevic — 2024-11-09

This study investigates the synergy between the virtual and real-world economies through e-commerce, where seller reputation is critical in guiding consumer decisions. As traditional businesses shift towards online retail, user reviews become essential, offering feedback to both sellers and potential buyers. Sentiment analysis through machine learning (ML) techniques presents significant advantages for consumers and retailers alike. This research proposes a novel approach combining bidirectional encoder representations from transformers (BERT) embeddings with an optimized XGBoost classification model to enhance sentiment analysis performance. A modified metaheuristic algorithm, derived from the firefly algorithm (FA), is introduced to optimize the model. Testing on publicly available datasets demonstrates that models optimized by this algorithm achieved a peak accuracy of .881336. Further statistical analyses substantiate these improvements, and SHAP interpretation on the best-performing model identifies key features impacting model predictions, shedding light on factors driving customer sentiment insights.

Research on Few-Shot Defect Detection Algorithm Based on Federated Learning

Yufeng Xiong — 2024-11-17

The algorithm based on deep learning has been widely used in defect detection in all walks of life, but the performance of the deep learning model depends mainly on rich annotation data. However, in the actual scene, obtaining large-scale, high-quality data to ensure users' privacy and safety is challenging, which limits its further promotion in specific application fields. To solve this problem, we propose a federated few-shot defect detection framework, which uses the privacy protection of the federated framework to jointly train independent few-shot tasks distributed on different clients to obtain a few-shot model that can quickly adapt to new tasks with limited data. We have done many experiments to evaluate our framework's effectiveness, and the results show that our framework is superior to the baseline and achieves the same performance as the model trained with a lot of data.

Dynamics Modelling and Adaptive Identification: Towards Improved Human-Robot Interaction in Collaborative Systems

Saixuan Chen — 2024-11-23

This article presents a novel approach to enhancing human-robot collaboration and safety through advanced dynamic modelling and adaptive identification techniques. We introduce a comprehensive methodology that integrates motion trajectory design with real-time torque detection, addressing the critical limitations of conventional systems that rely on costly joint torque sensors. By simultaneously identifying friction forces in an integrated joint and a simplified two-bar mechanism, our approach leverages existing kinematic and dynamic models to achieve precise dynamic parameter identification. The proposed method significantly advances the fields of drag-teaching and collision detection by eliminating the need for force sensors, thus making it more feasible for mass-produced robotic systems. Our findings demonstrate that accurate dynamic modelling is essential for effective zero-force control, particularly in high-speed drag-teaching scenarios, where inertia and friction present substantial challenges. Experimental validation confirms the efficacy of our dynamic feed-forward controller design and the adaptability of drag-teaching parameters, leading to improved operational flexibility and safety in collaborative environments. This research contributes a critical framework for future developments in intelligent robotic systems, providing a robust basis for integrating advanced human-robot interactions in industrial applications.