Prefacedoi: 10.1088/1742-6596/3022/1/011001pmid: N/A
2025 9th International Conference on Artificial Intelligence, Automation and Control Technologies (AIACT 2025), was successfully held in Sapporo, Japan from February 17 to 21, 2025, which was organized by Hong Kong Society of Mechanical Engineers(HKSME) and Shanghai Jiao Tong University, supported by Norwegian University of Science and Technology.Considering that some participants could not attend in person, the conference was adjusted as hybrid conference, as a combination of on-line and off-line conference.The conference accepted 33 papers, including countries like Malaysia, Japan, China, India, Thailand, Australia, Singapore,Canada, etc. Four renowned speakers delivered speeches about their latest research. They are Prof. Edwin K. P. Chong from Colorado State University, USA; Prof. Shugen Ma from The Hong Kong University of Science and Technology, China; Prof. Graziano Chesi from The University of Hong Kong, HKSAR,China and Prof. Haibin DUAN from Beihang University,China, delivered excellent speeches, sharing their latest and insightful research ideas. The conference also includes 3 technical sessions and one poster session. Each presenter was given 10-15 minutes to deliver their presentation, including 2 minutes Q&A. Two awards, one best oral presentation award and one best poster presentation award were selected by the end of conference. There is a lively discussion at the conference, which promotes academic exchange, which makes AIACT 2025 an effective communication platform for all the participants all over the world.List of COMMITTEES is available in this PDF.
Research on the Aviation Application of New High Strength Copper Aluminum Alloy Composite MaterialsQun, Sun
doi: 10.1088/1742-6596/3022/1/012017pmid: N/A
The aerospace industry demands conductor materials that combine high conductivity with lightweight, high-strength, and corrosion-resistant properties to support efficient electrical systems. This study explores the potential of copper-clad aluminum composite conductors, which combine copper’s conductivity with aluminum’s lightweight advantages, as a promising solution for aerospace applications. The research systematically evaluates the physical, chemical, and mechanical properties of copper-aluminum composite conductors, comparing them to pure copper and aluminum alloys in terms of weight, conductivity, tensile strength, and corrosion resistance. The results demonstrate that copper-aluminum composites offer a significant weight reduction without compromising electrical performance, making them suitable for use in aircraft wiring systems. Furthermore, the study highlights the technical challenges involved in manufacturing these composite materials, including the complexities of production and the need for new testing methods for compatibility with existing aerospace connectors. The study concludes that copper-clad aluminum conductors hold substantial promise for reducing aircraft weight, enhancing efficiency, and contributing to the development of cost-effective, high-performance aerospace materials.
Cascaded Multispectral Image Fusion with Saliency-Weighted Edge FeaturesGuo, Zhendong; Dong, Na; Wang, Yuwei; Mai, Xiaoming
doi: 10.1088/1742-6596/3022/1/012002pmid: N/A
In recent years, UAV inspection technology has been widely applied in infrastructure monitoring and power line inspection. However, significant differences between infrared and visible light images pose challenges for high-precision image registration. To address these challenges, a novel UAV multispectral image registration method based on saliency-weighted edge features and multi-feature cascade matching has been proposed. During feature extraction and description, this method uses a saliency-weighted grayscale window for edge extraction, combined with a multi-scale potential Harris corner selection algorithm to extract significant feature points. These edge features are then described using an eight-direction equal-area sector descriptor. In the feature matching phase, a cascaded matching framework is employed. It comprises an adaptive NNDR pre-screening based on domain priors to filter initial matches, followed by multi-feature fusion matching for fine-grained screening. A final geometric consistency check using the FSC algorithm is applied to effectively reduce the probability of mismatches. Experimental results demonstrate that this algorithm achieves an average high-precision matching rate of 90% on 97 pairs of infrared and visible light power equipment images provided by FLIR, attaining sub-pixel level accuracy. This performance significantly surpasses that of classic algorithms such as SIFT and its derivatives, SURF and ORB.
RT-DETR-Pothole: Lightweight Real-Time Detection Transformers for Improved Road Pothole DetectionFairuz Mat Radzi, Siti; Amiruddin Abd Rahman, Mohd; Luqman Arif Bin Mohamad, Muhammad
doi: 10.1088/1742-6596/3022/1/012003pmid: N/A
Assessment of on-time road condition is crucial for ensuring the safety of the motorist. One of the recent approaches to detecting road potholes is to analyze images captured from an unmanned aerial vehicle (UAV). Although the traditional deep learning model could perform accurate detection during offline analysis, there is still a limitation of the available algorithms that could perform real-time evaluation. Therefore, this study proposes a lightweight transformer algorithm, the real-time detection transformer (RT-DETR), for online evaluation of road pothole images. The models were tested in practical deployment scenarios and compared with several other object detection models, such as Faster RCNN- SqueezeNet, YOLOv8x, YOLOv9e, YOLOv10x, and YOLO11x. The results show that the RT-DETR- Pothole outperformed all other models in detection accuracy, achieving the highest mAP0.50 (0.834) and mAP0.50-0.95 (0.565), along with a high F1-Score (0.809), indicating superior precision and recall, and at the same time it could maintain low inference time. Overall, RT-DETR-Pothole is the most suitable model for real-time pothole detection, especially for detecting smaller, less visible potholes, with a reasonable inference time for pavement engineering applications.
TMOALO: A Sensor Node Load Balancing Approach for Power Internet of Things NetworksChen, Yanling; Zhu, Yuanjiao; Wei, Jingyi; Sun, Zheng; Jia, Dingyi; Zhang, Yao; Li, Jingsong; Li, Zegui; Li, Jingyun; Chen, Wenbin; Qu, Xin; Zhou, Jie
doi: 10.1088/1742-6596/3022/1/012008pmid: N/A
The improvement of service quality (QoS) and network lifetime extension are identified as urgent issues due to the widespread adoption of wireless sensor networks (WSNs) in Power Internet of Things (PIoT). During the secondary deployment phase, a typical multi-objective optimization challenge is constituted by how to maximize area coverage while the node movement distance is minimized. Therefore, a novel Tabu-based Multi-Objective Ant Lion Optimization algorithm (TMOALO) is proposed to address this challenge. Initially, a dynamic tabu search operator is designed to prevent the algorithm from being trapped in local optima, and thus the global search capability is enhanced. Subsequently, a layered elite preservation framework is introduced to ensure that high-quality solutions are effectively retained, whereby population diversity is increased. Furthermore, a nonlinear adaptive step-size control regulator is developed to optimize step adjustments, through which the stability and efficiency of the search process are improved. Experiments on a PIoT simulation platform demonstrate that the proposed TMOALO algorithm outperforms conventional methods like MALO, NSGA-II, and MOPSO, achieving an 11.005% increase in area coverage and a 4.118-meter reduction in sensor node movement. These results confirm its superiority in multi-objective optimization.
Deep Learning-driven Blind Spot Detection for Forklifts in Industrial EnvironmentsPornsing, Choosak; Karot, Thanathorn; Watanasungsuit, Arnat; Inta, Teerapat
doi: 10.1088/1742-6596/3022/1/012001pmid: N/A
In modern industrial settings, ensuring the safety of human workers coexisting with heavy machinery like forklifts is paramount. Traditional safety measures often fall short in addressing dynamic hazards, especially in blind spot areas where visibility is limited. This research introduces an advanced blind spot detection system utilizing deep learning techniques specifically designed for industrial environments. The system leverages the YOLOv8 architecture, enhanced through transfer learning, and optimized with model pruning and quantization techniques to achieve high accuracy and low latency, operating at 45 FPS on edge devices with a mean Average Precision (mAP) of 97.3%. The detection models are integrated with a real-time video processing pipeline using the RTSP protocol and OpenCV for spatial awareness. The system incorporates a novel attention mechanism to improve detection accuracy in challenging conditions, such as occlusions and varying lighting. Real-time alerts are provided via an integrated hardware system using ESP8266 microcontrollers and Node-RED for orchestration, ensuring immediate hazard notifications to operators. Extensive experiments in an automotive wheel manufacturing facility validate the system’s effectiveness, demonstrating an improvement of up to 5% in mAP compared to traditional methods. This work contributes to a scalable, efficient, and highly accurate solution for enhancing safety in high-risk industrial environments by addressing critical blind spot hazards in real time.
Peer Review Statementdoi: 10.1088/1742-6596/3022/1/011002pmid: N/A
All papers published in this volume have been reviewed through processes administered by the Editors. Reviews were conducted by expert referees to the professional and scientific standards expected of a proceedings journal published by IOP Publishing.• Type of peer review: Double Anonymous• Conference submission management system: Morressier• Number of submissions received: 62• Number of submissions sent for review: 46• Number of submissions accepted: 33• Acceptance Rate (Submissions Accepted / Submissions Received × 100): 53.2• Average number of reviews per paper: 2.727272727272727• Total number of reviewers involved: 26• Contact person for queries:Name: Lyn LeeEmail: [email protected]: Hong Kong Society of Mechanical Engineers - Conference Department
Research on Multi satellite Collaborative Task Planning TechnologyWang, J G; Qin, W T; Ran, D C; Wu, P; Wang, K; Liu, Y
doi: 10.1088/1742-6596/3022/1/012011pmid: N/A
As satellite technology advances and its applications expand, multi-satellite collaborative task planning has emerged as a key solution to address the growing demand for observation tasks amid limited satellite resources. By leveraging advanced planning algorithms and collaborative mechanisms, this approach optimizes resource allocation, enhances task execution efficiency, and reduces energy consumption and maintenance costs for satellites. Furthermore, multi-satellite collaborative task planning increases system robustness, allowing satellites to handle complex tasks and extreme conditions effectively. This article reviews recent research in this field, highlighting its significance in improving satellite system performance and meeting complex task requirements. The aim is to provide researchers with an overview of the current landscape and future directions in multi-satellite collaborative task planning, thus promoting the advancement of satellite technology and contributing to national security, economic growth, and scientific progress.
Enhancing Speech Recognition: Vowel Feature Extraction and Its Influence on Conformer Model EfficacyJang, Hyunsu; Kim, Jaekwang
doi: 10.1088/1742-6596/3022/1/012005pmid: N/A
Most deep-learning models of automatic speech recognition are complex and require large amounts of data for performance improvement. In the case of the XXL version of the Conformer-based model, performance is limited by the amount of data and a large parametric space (1 billion parameters), which requires significant computational resources. However, hardware and data are often inadequate for training this model; thus, it is necessary to identify performance improvement avenues notwithstanding limited resources. To this end, we propose a method for improved preprocessing, allowing to effectively extract the input data features. We present a method for strengthening the frequency of the vowel region by characterizing input voice data. The method was evaluated on the compact version of the Conformer model with 10 million parameters, the smallest among the existing Conformer models. Character error rates on the test clean dataset evaluation decreased by approximately 0.3% for the LibriSpeech 100-h-long dataset and 4.6% for the LibriSpeech 960-h-long dataset. In addition, for the LibriSpeech 100-h-long dataset, an improvement of 1.2% was obtained for the model in which the classification criterion was changed to sub-words, while an improvement of 0.6% was obtained for the sub-word and LibriSpeech 960-h-long dataset. These results show that input data preprocessing improves the performance of speech recognition models. The results of this study are reported only for the small Conformer model owing to hardware limitations, which in turn limits performance improvement. However, even with these limitations, the results strongly suggest that the model’s performance improved owing to the input data preprocessing.