Real-time System Research Articles

Efficient Visibility Approximation for Game AI using Neural Omnidirectional Distance Fields

Visibility information is critical in game AI applications, but the computational cost of raycasting-based methods poses a challenge for real-time systems. To address this challenge, we propose a novel method that represents a partitioned game scene as neural Omnidirectional Distance Fields (ODFs), allowing scalable and efficient visibility approximation between positions without raycasting. For each position of interest, we map its omnidirectional distance data from the spherical surface onto a UV plane. We then use multi-resolution grids and bilinearly interpolated features to encode directions. This allows us to use a compact multi-layer perceptron (MLP) to reconstruct the high-frequency directional distance data at these positions, ensuring fast inference speed. We demonstrate the effectiveness of our method through offline experiments and in-game evaluation. For in-game evaluation, we conduct a side-by-side comparison with raycasting-based visibility tests in three different scenes. Using a compact MLP (128 neurons and 2 layers), our method achieves an average cold start speedup of 9.35 times and warm start speedup of 4.8 times across these scenes. In addition, unlike the raycasting-based method, whose evaluation time is affected by the characteristics of the scenes, our method's evaluation time remains constant.

Research on a Real-Time Control System for Discrete Factories Based on Digital Twin Technology

Gear factories are most typical discrete manufacturing factories. Many gear factories are striving to explore how to achieve intelligent manufacturing in order to improve efficiency and reduce costs. Digital twin technology is currently one of the most reliable ways to achieve intelligent manufacturing. This article aims to establish a real-time control system in order to promote intelligent manufacturing for discrete manufacturing factories. Firstly, a model for a digital twin gear factory is put forward based on the characteristics of gear factories, and the composition of a real-time control system for gear factories is clarified. Then, a human–computer interaction architecture for the real-time control system is proposed. The real-time control system consists of three parts as follows: a monitoring module, a virtualizing module, and a controlling module. At work, it appears as a kind of human–machine interaction form with the three following interfaces: a monitoring window, a virtualizing window, and a controlling window. Finally, a gear factory, which is specialized in manufacturing the intermediate shaft dual gear of a new energy vehicle gearbox, develops a set of software for the real-time control system. The prototype software is obtained through some development activities such as 3D MAX and WebGL virtualization modeling and OPC UA and REST communication design.

The use of intelligent real-time image acquisition system using data mining technology in news acquisition and editing practice

Aiming to address the timely dissemination of news information, this work explores the clever utilization of data mining (DM) technology and deep learning (DL) algorithms to construct an intelligent real-time news image acquisition system to meet the urgency of news dissemination needs. First, this work introduces an intelligent real-time news image acquisition system and provides a detailed analysis of its principles and advantages. Throughout this process, the crucial role of DM technology in news image classification and automation is emphasized, especially in dealing with rapidly evolving news events. Next, the work establishes an intelligent real-time news image acquisition model based on DL algorithms, which integrates the essence of DM technology. Through this fusion, the research objective is to enhance the performance of the news image acquisition system to achieve higher real-time and accuracy, which is vital for the swift delivery of news information. Finally, this work investigates the application of the intelligent news image acquisition system in network communication to ensure its adaptability to various network communication scenarios while maintaining accuracy and real-time capabilities. The research results demonstrate that the application of DM technology in combination with DL algorithms can effectively meet the practical needs of the news industry, enhancing the automation of news image processing and enabling faster information delivery to the audience. Notably, the AlexNet model employed performs exceptionally well, achieving recognition rates of up to 99.6% after data augmentation or equalization processing, with an accuracy of 90.9% and a high specificity of 93.38%. This indicates outstanding overall classification accuracy and negative class accuracy, even when distinguishing between news and non-news scenarios. These results clearly underscore the connection between DM technology and news acquisition and editing practices, and emphasize its potential to improve the efficiency and accuracy of real-time information dissemination. The research’s contribution and innovation lie in the fusion of DM technology with DL algorithms to build an intelligent real-time news image acquisition system. This fusion enhances the automation and classification performance of news images, thereby improving the real-time and accuracy of news information. Furthermore, the work strongly emphasizes improving the real-time and accuracy of the news image acquisition system to ensure the swift delivery of information, which is of utmost importance in rapidly changing news events.

Online identification and feed-forward compensation of nonlinear friction in servo system based on RBF neural network model

In this paper, an online identification and compensation method of nonlinear friction based on radial basis function (RBF) neural network model is proposed for the influence of nonlinear friction on machining accuracy in the low speed process of servo feed system of CNC machine tools. First, a three-layer single-input-output RBF neural network model is established for describing the nonlinear friction of servo feeding system. Second, the neural network online learning algorithm is improved based on adaptive gain, which improves the stability and accuracy of the algorithm. Finally, experiments were carried out on a three-axis milling machine to compensate the friction in the servo feed system in real time based on the online identification results. The results show that the method can effectively improve the online identification accuracy and convergence rate, and effectively improved the low-speed performance of the servo feed system.

TEDNet: Twin Encoder Decoder Neural Network for 2D Camera and LiDAR Road Detection

Abstract Robust road surface estimation is required for autonomous ground vehicles to navigate safely. Despite it becoming one of the main targets for autonomous mobility researchers in recent years, it is still an open problem in which cameras and LiDAR sensors have demonstrated to be adequate to predict the position, size and shape of the road a vehicle is driving on in different environments. In this work, a novel Convolutional Neural Network model is proposed for the accurate estimation of the roadway surface. Furthermore, an ablation study has been conducted to investigate how different encoding strategies affect model performance, testing 6 slightly different neural network architectures. Our model is based on the use of a Twin Encoder–Decoder Neural Network (TEDNet) for independent camera and LiDAR feature extraction and has been trained and evaluated on the Kitti–Road dataset. Bird’s Eye View projections of the camera and LiDAR data are used in this model to perform semantic segmentation on whether each pixel belongs to the road surface. The proposed method performs among other state-of-the-art methods and operates at the same frame rate as the LiDAR and cameras, so it is adequate for its use in real-time applications.

Improvement of butt-fusion welding procedure and performance evaluation method for thick-walled bimodal polyethylene pipes

High-density polyethylene (HDPE) pipes have been continuously improved by the progress made by the resin industry and pipe manufacturers. Together with the advanced technology of welding equipment in terms of structure reliability and automations, HDPE pipes have been gradually applied to higher pressure level, bigger diameter, and thicker walls in operations which have become the preferred piping system for nuclear safety application in transporting cooling seawater. There are substantial differences in parameter settings when applying heat and force during different welding procedures, as well as the complicated evaluation methods for mechanical properties, which causes communication barriers between engineers and researchers. Thus, it is necessary to comprehensively study the welding procedures and evaluation methods to standardize the evaluation protocols in the industry. In this paper, based on the assumptions of temperature effect and molecular chain diffusion dynamics at the welding interface, the optimization of the welding procedure is proposed. Some well-established evaluation methods have been used for performance assessment and their feasibilities are discussed. The idea of creating crack along the weld interface is adopted as new method to evaluate the material ability of fracture resistance. Specimens with unilateral notch and weld interface crack are proposed and subjected for quasi-static axial tensile test. The fracture stress and fracture energy at the welding interface are used to differentiate the quality of the joint from different welding procedures. The fracture morphologies of the weld interface are analyzed. A complete picture is drawn for the mechanical fracture resistance of the weld interface. The method used in this paper can be corelated to the pipeline design conditions, therefore could be applied to the real-time engineering applications. Moreover, the evaluation method proposed in this paper is simple, reliable, and easy to implement.

Real-time detection of surface cracking defects for large-sized stamped parts

This study presents a framework for the real-time detection of surface cracking in large-sized stamped metal parts. The framework aims to address the challenges of low detection efficiency and high error rates associated with manual cracking detection. Within this framework, a novel network, SNF-YOLOv8, is proposed to efficiently detect cracking while ensuring that the detection speed matches the production speed. The network incorporates a convolutional spatial-to-depth module to enhance the detection of small-sized cracking and mitigate surface interference during inspections. Furthermore, a visual self-attention mechanism is introduced to improve feature extraction. A combination of standard convolutional and depth-wise separable convolutional layers in the neck network enhances speed without compromising accuracy. Experimental validation conducted using a dataset from actual production lines, in collaboration with a multi-national corporation, demonstrates that SNF-YOLOv8 achieves an average precision of 85.2% at a detection speed of 164 frames per second. The framework achieves an accuracy rate of 98.8% in detecting large-sized cracking and 96.4% in detecting small-sized cracking, meeting the requirements for high-precision and real-time detection applications.

The development and impact of an app for a smart drug interaction reminder system.

Improved access to media and medical knowledge has elicited stronger public health awareness. This study developed a smart drug interaction reminder system for patients to increase knowledge and reduce nurse workload. This study used a single-group pre-test/post-test design and applied mining techniques to analyze the weight and probability of interaction among various medicines. Data were collected from 258 participants at a teaching hospital in northern Taiwan using convenience sampling. An app was used to give patients real-time feedback to obtain access to information and remind them of their health issues. In addition to guiding the patients on medications, this app measured the nurses' work satisfaction and patients' knowledge of drug interaction. The results indicate that using information technology products to assist the app's real-time feedback system promoted nurses' work satisfaction, improved their health education skills, and helped patients to better understand drug interactions. Using information technology to provide patients with real-time inquiring functions has a significant effect on nurses' load reduction. Thus, smart drug interaction reminder system apps can be considered suitable nursing health education tools and the SDINRS app can be integrated into quantitative structure-activity relationship intelligence in the future.

Front-end speech processing system with SVM algorithm in rail passenger flow management

In order to achieve the goal of dynamically adjusting daily passenger flow to effectively control the overall efficiency of the transportation system, this study constructs a real-time monitoring and prediction system for subway passenger flow based on front-end voice processing technology and support vector machine models. The study first conducted a railway passenger flow analysis, and then used a support vector machine model to construct a preliminary prediction system. In order to achieve global optimization, the study also introduced particle swarm optimization algorithm to construct an optimization prediction model based on PSO-SVM. The results show that the proposed PSO-SVM method has undergone 48 iterations of training, and the predicted values closely match the actual passenger flow curve. The maximum RE error is 2%, and the overall prediction error is 98%. The decision coefficient of PSO-SVM is 0.998932. Therefore, this indicates that it has high performance and feasibility in predicting and controlling passenger flow during peak hours of urban rail transit.

An efficient method for modeling and evaluating the bench terrain of open-pit mines

In order to quantitatively analyze the roughness of the bench floor during open-pit mine blasting, this study proposes a real-time measuring method for the three-dimensional terrain of the bench floor during the excavation process. Real-time monitoring is conducted at the boundary and discrete internal points of the workbench floor during electric shovel operation, utilizing real-time kinematic global navigation satellite system (RTK-GNSS) positioning technology. An improved convex hull algorithm is introduced to automatically extract the optimal boundary of discrete point clouds based on their spatial distribution characteristics. This study establishes a digital elevation model (DEM) using five interpolation algorithms for 3D terrain visualization simulation. Through cross-validation, a comparative analysis of the DEM accuracy, the simulation results of the ordinary kriging interpolation algorithm were found to be optimized. The optimized interpolation algorithm is applied to simulate the 3D terrain in the Dexing open-pit copper mine, and the relevant terrain parameters were calculated. This dataset can serve as a precise foundation for the real-time path planning of elevation blasting design and ground leveling operations.

Sub-microscopic Plasmodium falciparum infections and multiple drug resistant single nucleotide polymorphic alleles in pregnant women from southwestern Nigeria

ObjectivesThe study evaluated sub-microscopic malaria infections in pregnancy using two malaria Rapid Diagnostic Tests (mRDTs), microscopy and RT-PCR and characterized Plasmodium falciparum dihydrofolate reductase (Pfdhfr) and Plasmodium falciparum dihydropteroate synthase (Pfdhps) drug resistant markers in positive samples.MethodsThis was a cross sectional survey of 121 pregnant women. Participants were finger pricked, blood drops were collected for rapid diagnosis with P. falciparum histidine-rich protein 11 rapid diagnostic test kit and the ultra-sensitive Alere Pf malaria RDT, Blood smears for microscopy and dried blood spots on Whatman filter paper for molecular analysis were made. Real time PCR targeting the var acidic terminal sequence (varATS) gene of P. falciparum was carried out on a CFX 96 real time system thermocycler (BioRad) in discriminating malaria infections. For each run, laboratory strain of P. falciparum 3D7 and nuclease free water were used as positive and negative controls respectively. Additionally, High resolution melt analyses was employed for genotyping of the different drug resistance markers.ResultsOut of one hundred and twenty-one pregnant women sampled, the SD Bioline™ Malaria Ag P.f HRP2-based malaria rapid diagnostic test (mRDT) detected eight (0.06%) cases, the ultra-sensitive Alere™ malaria Ag P.f rapid diagnostic test mRDT had similar outcome in the same samples as detected by the HRP2-based mRDT. Microscopy and RT-PCR confirmed four out of the eight infections detected by both rapid diagnostic tests as true positive and RT-PCR further detected three false negative samples by the two mRDTs providing a sub-microscopic malaria prevalence of 3.3%. Single nucleotide polymorphism in Pfdhps gene associated with sulphadoxine resistance revealed the presence of S613 mutant genotypes in three of the seven positive isolates and isolates with mixed wild/mutant genotype at codon A613S. Furthermore, four mixed genotypes at the A581G codon were also recorded while the other Pfdhps codons (A436G, A437G and K540E) showed the presence of wild type alleles. In the Pfdhfr gene, there were mutations in 28.6%, 28.6%, and 85.7% at the I51, R59 and N108 codons respectively. Mixed wild and mutant type genotypes were also observed in 28.6% each of the N51I, and C59R codons. For the Pfcrt, two haplotypes CVMNK and CVIET were observed. The SVMNT was altogether absent. Triple mutant CVIET 1(14.3%) and triple mutant + wild genotype CVIET + CVMNK 1(14.3%) were observed. The Pfmdr1 haplotypes were single mutants YYND 1(14.3%); NFND 1(14.3%) and double mutants YFND 4(57.1%); YYDD 1(14.3%).

Schedulability Analysis in Fixed-Priority Real-Time Multicore Systems with Contention

In the scheduling of hard real-time systems on multicore platforms, significant unpredictability arises from interference caused by shared hardware resources. The objective of this paper is to offer a schedulability analysis for such systems by assuming a general model that introduces interference as a time parameter for each task. The analysis assumes constrained deadlines and is provided for fixed priorities. It is based on worst-case response time analysis, which exists in the literature for monocore systems. We demonstrate that the worst-case response time is an upper bound, and we evaluate our proposal with synthetic loads and execution on a real platform.

Highway Deformation Monitoring by Multiple InSAR Technology

Addressing the challenge of large-scale uneven deformation and the complexities of monitoring road conditions, this study focuses on a segment of the G15 Coastal Highway in Jiangsu Province. It employs PS-InSAR, SBAS-InSAR, and DS-InSAR techniques to comprehensively observe deformation. Analysis of 73 image datasets spanning 2018 to 2021 enables separate derivation of deformation data using distinct InSAR methodologies. Results are then interpreted alongside geological and geomorphological features. Findings indicate widespread deformation along the G15 Coastal Highway, notably significant settlement near Guanyun North Hub and uplift near Guhe Bridge. Maximum deformation rates exceeding 10 mm/year are observed in adjacent areas by all three techniques. To assess data consistency across techniques, identical observation points are identified, and correlation and difference analyses are conducted using statistical software. Results reveal a high correlation between the monitoring outcomes of the three techniques, with an average observation difference of less than 2 mm/year. This underscores the feasibility of employing a combination of these InSAR techniques for road deformation monitoring, offering a reliable approach for establishing real-time monitoring systems and serving as a foundation for ongoing road health assessments.

MnO/Mn2O3 Aerogels as Effective Materials for Supercapacitor Applications

Mixed-oxide transition-metal aerogels (AGLs), particularly manganese-based AGLs, have attracted considerable interest over the past decade owing to their extraordinary properties, including high porosity, good surface area, and ultralow density. To develop easy and lightweight materials for the ever-increasing energy storage demands of the near future, we designed a novel Mn-based electrode material to meet these rising requirements. MnO/Mn2O3 AGLs were synthesized using a novel borohydride hydrolysis method and then annealed at 200, 400, and 550 °C. The as-synthesized AGLs yielded flower-like network structures, but their porosity increased with increasing temperatures, to a high temperature of 400 °C. This increased porosity and network structure facilitate a high capacitance. A supercapacitor (SC) constructed with the three-electrode material yielded 230 F/g for the MnAGL@400 sample, followed by yields from the MnAGL@200 and MnAGL@550 electrodes. Furthermore, the device constructed with MnAGL@400 exhibited an energy density of 9.8 Wh/kg and a power density of ~16,500 W/kg at a current density of 20 A/g. The real-time applicability of the AGL was demonstrated by engineering a two-electrode device employing MnAGL@400 as the positive electrode, which exhibited 97% capacity retention and 109% Coulombic efficiency over 20,000 cycles.

A Machine Learning-Based Tropospheric Prediction Approach for High-Precision Real-Time GNSS Positioning

For high-precision positioning applications, various GNSS errors need to be mitigated, including the tropospheric error, which remains a significant error source as it can reach up to a few meters. Although some commercial GNSS correction data providers, such as the Quasi-Zenith Satellite System (QZSS) Centimeter Level Augmentation Service (CLAS), have developed real-time precise regional troposphere products, the service is available only in limited regional areas. The International GNSS Service (IGS) has provided precise troposphere correction data in TRO format post-mission, but its long latency of 1 to 2 weeks makes it unable to support real-time applications. In this work, a real-time troposphere prediction method based on the IGS post-processing products was developed using machine learning techniques to eliminate the long latency problem. The test results from tropospheric predictions over a year using the proposed method indicate that the new method can achieve a prediction accuracy (RMSE) of 2 cm, making it suitable for real-time applications.

Towards real-time practical image compression with lightweight attention

The rate–distortion (RD) performance of learning-based image compression (LIC) has already surpassed that of traditional Versatile Video Coding (VVC) intra-coding. However, the gain of the compression efficiency is at the cost of high computational complexity, which is prohibitively expensive and becomes a bottleneck for practical applications. To this end, we proposed an efficient LIC method with lightweight designs for real-time practical applications, which achieves a better trade-off between compression performance and computational complexity. Specifically, residual connected lightweight attention units (RLAUs) are stacked for feature extraction, which achieves effective global spatial information embedding while keeping relatively low complexity. Meanwhile, a trainable channel-gained adaptive module (CGAM) is introduced into the nonlinear transform network and multi-stage context model. This module re-distributes the importance of different channels, further improving compression efficiency. The experimental results demonstrate that the proposed method achieves better compression efficiency than VVC intra-coding. Furthermore, compared to other state-of-the-art LIC approaches, the proposed method significantly reduces the coding time while maintaining comparable coding efficiency. The source codes are available at https://github.com/llsurreal919/LightweightLIC.

Developing a flood monitoring system by utilizing real-time satellite rainfall estimates and water energy budget-based rainfall-runoff inundation model in West Africa

Abstract. The African regions have been affected by frequent flood-related disasters. To support decision-makers in flood disaster management, a real-time flood monitoring system was established in the Niger and Volta basins under the Water disaster platform to enhance climate resilience in Africa (WADiRE-Africa). To bridge a gap between data unavailability and inaccessibility, the near-real-time satellite rainfall data (GSMaP-NRT) was evaluated and found to have a very low correlation (<0.4) with gauge data at daily scales and a higher correlation (>0.7) at monthly scales. The statistically bias-corrected GSMaP-NRT data was used in the Water and Energy Budget-based Rainfall-Runoff-Inundation (WEB-RRI) model to generate hydrological responses of the basins. As the simulated river discharges were overestimated, an additional correction method was employed for the GSMaP-NRT data to match well with observed discharges. The prototype system provides real-time flood information in both basins.

Correction: Real-time Machine Vision System for the Visually Impaired

Correction: Real-time Machine Vision System for the Visually Impaired

An efficient visual servo tracker for herd monitoring by UAV

It is a challenging and meaningful task to carry out UAV-based livestock monitoring in high-altitude (more than 4500 m on average) and cold regions (annual average – 4 °C) on the Qinghai Tibet Plateau. The purpose of artificial intelligence (AI) is to execute automated tasks and to solve practical problems in actual applications by combining the software technology with the hardware carrier to create integrated advanced devices. Only in this way, the maximum value of AI could be realized. In this paper, a real-time tracking system with dynamic target tracking ability is proposed. It is developed based on the tracking-by-detection architecture using YOLOv7 and Deep SORT algorithms for target detection and tracking, respectively. In response to the problems encountered in the tracking process of complex and dense scenes, our work (1) Uses optical flow to compensate the Kalman filter, to solve the problem of mismatch between the target bounding box predicted by the Kalman filter (KF) and the input when the target detection in the current frame is complex, thereby improving the prediction accuracy; (2) Using a low confidence trajectory filtering method to reduce false positive trajectories generated by Deep SORT, thereby mitigating the impact of unreliable detection on target tracking. (3) A visual servo controller has been designed for the Unmanned Aerial Vehicle (UAV) to reduce the impact of rapid movement on tracking and ensure that the target is always within the field of view of the UAV camera, thereby achieving automatic tracking tasks. Finally, the system was tested using Tibetan yaks on the Qinghai Tibet Plateau as tracking targets, and the results showed that the system has real-time multi tracking ability and ideal visual servo effect in complex and dense scenes.

Enhancing Quality Control: Image-Based Quantification of Carbides and Defect Remediation in Binder Jetting Additive Manufacturing

While binder jetting (BJ) additive manufacturing (AM) holds considerable promise for industrial applications, defects often compromise part quality. This study addresses these challenges by investigating binding mechanisms and analyzing common defects, proposing tailored solutions to mitigate them. Emphasizing defect identification for effective quality control in BJ-AM, this research offers strategies for in-process rectification and post-process evaluation to elevate part quality. It shows how to successfully process metallic parts with complex geometries while maintaining consistent material properties. Furthermore, the paper explores the microstructure of AISI M2 tool steel, utilizing advanced image processing techniques like digital image analysis and SEM images to evaluate carbide distribution. The results show that M2 tool steel has a high proportion of M6C carbides, with furnace-cooled samples ranging from ~2.4% to 7.1% and MC carbides from ~0.4% to 9.4%. M6C carbides ranged from ~2.6% to 3.8% in air-cooled samples, while water-cooled samples peaked at ~8.52%. Sintering conditions also affected shrinkage, with furnace-cooled samples showing the lowest rates (1.7 ± 0.4% to 5 ± 0.4%) and water-cooled samples showing the highest (2 ± 0.4% to 14.1 ± 0.4%). The study recommends real-time defect detection systems with autonomous corrective capabilities to improve the quality and performance of BJ-AM components.