Dynamic Information Environment Research Articles

Integration of eye-tracking and object detection in a deep learning system for quality inspection analysis

Abstract During quality inspection in manufacturing, the gaze of a worker provides pivotal information for identifying surface defects of a product. However, it is challenging to digitize the gaze information of workers in a dynamic environment where the positions and postures of the products and workers are not fixed. A robust, deep learning-based system, ISGOD (Integrated System with worker’s Gaze and Object Detection), is proposed, which analyzes data to determine which part of the object is observed by integrating object detection and eye-tracking information in dynamic environments. The ISGOD employs a six-dimensional pose estimation algorithm for object detection, considering the location, orientation, and rotation of the object. Eye-tracking data were obtained from Tobii Glasses, which enable real-time video transmission and eye-movement tracking. A latency reduction method is proposed to overcome the time delays between object detection and eye-tracking information. Three evaluation indices, namely, gaze score, accuracy score, and concentration index are suggested for comprehensive analysis. Two experiments were conducted: a robustness test to confirm the suitability for real-time object detection and eye-tracking, and a trend test to analyze the difference in gaze movement between experts and novices. In the future, the proposed method and system can transfer the expertise of experts to enhance defect detection efficiency significantly.

A Reliable Physical Layer Key Generation Scheme Based on RSS and LSTM Network in VANET

With the rapid development of information and communication technology, Vehicular AD hoc Networks (VANET) has attracted more and more attention. In order to provide traffic safety services, vehicles frequently share information related to road traffic, such as vehicle motion data and traffic flow data. Reliable key generation is the basis of VANET security system construction. At present, most key generation schemes rely on a trusted third-party, so there are security risks. Traditional key agreement protocols have high overhead, and is not suitable for the latency-sensitive requirements of VANET. The physical layer security technology extracts the fingerprint of the wireless channel and the identity of the device, and generates the key quickly and in real time without the third party distribution. A physical layer key generation scheme based on Received Signal Strength (RSS) is proposed to realize Vehicle-to-Vehicle (V2V) secure communications. Firstly, a network model based on Long Short-Term Memory (LSTM) network and Kalman Filtering is proposed to effectively enhance the reciprocity of physical layer information in dynamic environment. Secondly, a lossless quantization scheme is proposed, which achieves a lower bit disagreement rate and a higher bit generation rate. Thirdly, the inconsistent bits are corrected by fuzzy extractor, the confidentiality of the key exchange process is enhanced by zero knowledge proof, and the security of the key is improved by using hash function for privacy amplification. Finally, the experimental results show that the proposed scheme has great improvements in data correlation, bit disagreement rate, bit generation rate and bit distribution randomness.

IMPLEMENTATION OF THE PRINCIPLES OF SCIENTIFIC AND RESEARCH ACTIVITY OF FUTURE TECHNOLOGY TEACHERS IN THE EDUCATIONAL PROCESS OF HIGHER EDUCATION INSTITUTIONS

The article examines the importance and methods of introducing the principles of scientific research into the educational process of future technology teachers, emphasizing the importance of forming scientific consciousness, developing critical thinking, and the ability to innovatively apply knowledge in practical activities. Modern approaches to the integration of scientific research into the educational process of higher education institutions are analyzed, including the usage of active and interactive teaching methods, and the involvement of students in real research projects that contribute to the development of scientific and research skills.
 The key role of the implementation of scientific and research principles in the educational process of training future technology teachers is emphasized, and the importance of this aspect for the development of the necessary competencies in a dynamic information and technology environment is emphasized. The role of active and interactive teaching methods in involving students in a deep understanding of scientific concepts and practical application of knowledge is defined. The value of elective disciplines for the development of scientific and research abilities and skills, which determines the preparation of future teachers for independent research work, is clarified. An overview of opportunities for applicants to participate in scientific projects and publish their work is provided, which contributes to professional growth and scientific careers. The need to provide access to modern resources and improve the qualifications of teachers for the effective implementation of scientific and research activities in education is emphasized. A comprehensive approach to reforming curricula and teaching methods is proposed, which is crucial for the formation of highly qualified technology teachers ready for innovative activities and scientific research. Such systematic preparation for scientific research activity will contribute to the formation of future specialists’ readiness for innovative, creative, and critical professional activity.
 A conclusion has made about the need for further development of methods of integration of scientific research activities into educational processes to ensure a high level of training future technology teachers, taking into account the dynamically changing requirements of the modern educational and technological environment.

Path planning method for Camellia oleifera forest trenching operation based on human-robot collaboration

In order to realize the independent and safe operation of trenching robot in Camellia oleifera forest, operation path planning is the key link. Aiming at the problem of trenching operation path planning in complex dynamic Camellia oleifera forest environment, a trenching operation path planning method of Camellia oleifera forest based on human-robot collaboration was proposed. In order to meet the requirements of trenching operation in Camellia oleifera forest, the ring trenching operation path was planned around Camellia oleifera tree; Aiming at the unknown obstacles in the dynamic Camellia oleifera forest environment, the digital twin technology was used to map the dynamic Camellia oleifera forest environment into the twin space in real time, assist the operator to recognize the dynamic Camellia oleifera forest environment information and issue decision instructions. According to the man-machine collaborative decision-making instructions and dynamic environmental information, the twin environment model of Camellia oleifera forest was updated in real-time, and the collision detection method combining virtual and real was adopted to avoid collision and re-plan the trenching operation path. The experimental results show that the repetition rate of the trenching operation path of the Camellia oleifera forest planned by this method is reduced by 1.64 percentage points, the slope standard deviation is reduced by 40.365 %, and the operation path is more stable; In addition, the re-planning of the operation path of the trenching robot based on digital twin under human-robot cooperation prevents collision with the dynamic obstacles of the Camellia oleifera forest, which provides a new idea for the intelligentization of the trenching operation of the Camellia oleifera forest.

SIA-SLAM: a robust visual SLAM associated with semantic information in dynamic environments

SIA-SLAM: a robust visual SLAM associated with semantic information in dynamic environments

IndusSynthe: Synthetic data using human-machine intelligence hybrid for enhanced industrial surface defect detection through self-updating with multi-view filtering

Artificial intelligence (AI) has made significant strides in automating defect detection for industrial products. However, its real-time application in manufacturing, especially with limited data, is not extensively explored. Current AI systems predominantly focus on post hoc analysis, encountering challenges of data scarcity, diversity, and scalability of unknown data. In this study, we proposed a novel method, IndusSynthe, which is a zero-data detection method with self-updating using human–machine intelligence hybrid (HMIH) and multi-view filtering (MVF). Firstly, generate diverse data enriched with expert knowledge, achieving the model training in zero-data scenarios. Subsequently, an MVF strategy is employed as a self-updating learning mechanism, addressing challenge associated with declining model accuracy in dynamic information environments. In subsequent iterations, only this second phase is executed to ensure a consistent and stable self-updating process. Two emblematic industrial datasets were employed to validate the efficacy and feasibility of the proposed method. Experimental outcomes indicate that the method enables initial training in a zero-data environment by integrating human–machine knowledge, negating the need for extensive auxiliary data. Furthermore, it empowers machines to autonomously handle self-screening and updating in dynamic industrial settings, thereby establishing robust and efficient anomaly detection mechanisms. The proposed method expedites the development and assessment of an intelligent system tailored for industrial surface defect detection, it offers a comprehensive solution, from inception to application, for incorporating AI technologies in industrial manufacturing. We envisage that this method may inspire future research avenues and broaden its relevance to additional vision-centric tasks.

A Multirobot Cooperative Area Coverage Search Algorithm Based on Bioinspired Neural Network in Unknown Environments

This article proposes a novel approach based on a bioinspired neural network (BIN) for multirobot area coverage search in unknown environments. We obtain the dynamic environment information in the search process of the multirobot by combining the BIN model with the grid map. We, then, design a search objective function to guide robots to explore the task area by applying the distributed model prediction control policy. Moreover, we further improve the search efficiency through a collaborative decision-making mechanism among robots. Finally, we conduct simulation studies, which demonstrate the effectiveness and superiority of the proposed approach.

Pulsed turbidite and methane seep records in the north western South China Sea since the last glacial maximum

The small-scale event layers in the continental margin contain abundant dynamic environment information, and pose a challenge to the interpretation of continuous sedimentary records, giving geological significance to their accurate identify and possible genesis. Here, pulsed turbidite layers since the last glacial maximum (LGM) in a gravity core in the northwestern South China Sea (SCS) was analyzed to investigate the precisely identification, possible causes and the role of marine environmental change during the late Quaternary in formation of these small-scale event layers in the SCS. Eight potential pulsed turbidite layers, according to the petrographic characteristics, grain size parameters and element geochemistry, were identified. Meanwhile, indicators including total sulfur (TS)/total organic carbon (TOC) ratio, CaCO3 content, and chromium-reducible sulfur (CRS) revealed these horizons were mostly related to methane seep events. Constrained by foraminifera shells AMS14C results, these events were determined to have occurred from the LGM to early Holocene, Similar records in the northern and southern slopes suggests the universal occurrence of these small-scale layers in the SCS. The comprehensive analysis showed that the development of these event layers over the past 25 ka can be divided into three stages, 25-15.5 ka, 15.5-7 ka and 7 ka to present. Late Quaternary Ocean environment changes, especially sea level and bottom water temperature, controlled the occurrence of regional small-scale event layers in the SCS. The regional scale mechanism is that the pressure and temperature change affect the stability of hydrate and the methane seepage, and thus the strata stability. Corresponding to the lowest, the rapid increase and the highest levels of the sea level and bottom water temperature, the temporal evolution pattern of small-scale event layers in the SCS showed a highest, decreased and lowest frequency, respectively. The linkage between the late Quaternary marine environmental change and turbidite deposition through gas activities in this study can act as a useful reference for further understanding the continental margin sedimentary process.

Fedab: A Low-Latency Energy-Efficient Proactive Neighbor Discovery Protocol in MLDC-WSN

Mobile Low-Duty-Cycle Wireless Sensor Network (MLDC-WSN) is a new type of wireless multi-hop network composed of sensor nodes that have mobility and sleep for a long time in a wireless self-organizing manner. Neighbor discovery protocols are used for sensor nodes to discover neighbor nodes within their communication range and can provide neighbor information to other protocols. MLDC-WSN uses low duty cycle technology to reduce energy consumption, and captures dynamic environmental information through the movement of nodes. However, the random mobility and long dormancy of nodes lead to frequent changes in the network topology, and nodes need to consume a lot of energy and time to discover all their neighbors. The sensor nodes are often powered by energy-limited batteries, and battery replacement is difficult. How to implement efficient neighbor discovery with low-latency and low-energy is a challenge of current research. Different from the previous slot model that couples the listening and transmitting of beacons in one slot, this paper proposes a new discovery schedule based on the slot model that separates the listening and transmitting of beacons. The discovery schedule can guarantee unidirectional discovery. And the node proactively transmits the beacon to accelerate bidirectional discovery after unilaterally discovering its neighbor nodes. On this basis, a proactive asynchronous neighbor discovery protocol called Fedab is proposed in this paper. Ignoring beacon collisions and other realistic interference factors, Fedab can enable neighbor discovery within the theoretical worst-case discovery latency. Moreover, Fedab is extended to the scenario where nodes perform different duty cycles, i.e., asymmetric scenarios. After experimental simulations and numerical analysis, the Fedab protocol has better performance compared to other existing neighbor discovery protocols.

Discrimination of Oceanic Whitecaps Derived by Sea Surface Wind Using Sentinel‐2 MSI Images

AbstractOceanic whitecaps, as a manifestation of the wind‐wave breaking process and the medium of air‐sea exchange, play an important role in sea surface research. The observational data of oceanic whitecap coverage are generally in situ photographs or videos. The use of spaceborne optical sensors to capture oceanic whitecaps could provide more spatial distribution and dynamic marine environment information. In this study, more than 100 Sentinel‐2 Multi‐Spectral Instrument (MSI) images and corresponding synchronous buoy wind speed data were collected to investigate oceanic whitecaps. A curvature image processing method was employed to successfully identify oceanic whitecaps in MSI images and the whitecap coverage was estimated based on the whitecap recognition results. We established a power‐law model of MSI‐estimated whitecap coverage (W) and whitecap number density with buoy‐measured wind speed at a height of 10 m above the sea surface (U10). In addition, the differences between the MSI‐derived model and in situ measurement‐derived models which are mainly caused by the scale effect of optical remote sensing, are discussed. The whitecap‐wind model demonstrates that MSI images can improve the quantification of sea surface wind speed with high spatial resolution. MSI images were used to estimate regional high‐spatial resolution (4 km) wind speeds, and the merged results (0.25°) were verified with the hourly wind products of the fifth generation European ECMWF (Centre for Medium‐Range Weather Forecasts) atmospheric reanalysis (ERA5). Moreover, the influence of sunglint on whitecap recognition and whitecap coverage was also analyzed. The reflectance of whitecap‐affected pixels in MSI images increased obviously with wind speed compared with background seawater in weak sunglint reflection area. These results indicated that the parameters of oceanic whitecaps derived from high‐spatial resolution optical images can be used to estimate wind speed and provide a new approach for monitoring marine dynamic environments in the future.

An approach to improve argumentation-based epistemic planning with contextual preferences

Current approaches to argumentation-based planning represent an interesting proposal where defeasible argumentation is used as a practical mechanism suitable for reasoning with potentially contradictory information in dynamic environments. In many real-world planning scenarios, the development of formalisms allowing explicit preference specification over pieces of knowledge turns out to be an essential task—however, despite its importance, existing planning systems are not provided with the possibility of dynamically changing these preferences when a plan is being constructed. This paper presents an argumentation-based approach to deal with the handling of preferences when a plan is formulated; in particular, we propose using conditional expressions to select and change priorities regarding information upon which plans are constructed. Our aim is not to improve the efficiency of current planning systems, but to enhance the resulting plan itself by introducing an approach capable of representing and handling multiple preferences over defeasible knowledge. This approach will contribute to the strengthening of existing argumentation-based epistemic planning systems, providing a useful tool that the user could exploit. Finally, we also present a running-time analysis and several complexity results associated with our approach.

A collaborative optimization strategy for computing offloading and resource allocation based on multi-agent deep reinforcement learning

With the emergence of mobile edge computing (MEC), the edge cloud with certain computing power is deployed closer to the mobile device, which can well solve the computing and delay requirements of the mobile device. In 5G ultra-dense heterogeneous networks, where the macro base station (MBS) and multiple dense small base stations (SBS) are deployed in the region, the offloading decision faces multiple choices. In order to solve the problem of computing offloading and resource allocation in 5G ultra-dense heterogeneous networks, we propose a collaborative optimization strategy based on multi-agent deep reinforcement learning (MADRL). At each time, the mobile device only needs to make the optimal offloading decision according to its own historical offloading decision, the allocated bandwidth and computing resources at the past time, as well as the service response delay and energy consumption at the past time, without knowing other user information and dynamic network environment information. Simulation results show that the proposed collaborative optimization strategy is better than the other three baseline schemes in terms of service response delay and energy consumption performance.

MotiLearn: Contract-Based Incentive Mechanism for Heterogeneous Edge Collaborative Training

Due to the computation capacity improvement of edge devices and the popularization of artificial intelligence, there has been a dramatic increase in topic of edge intelligence. In order to motivate edge users to participate in task release and collaborative training, it is crucial to design a feasible incentive mechanism so that the task publisher (Principal) can maximize his interests while guaranteeing the income of co-trainers (Agent). In this paper, we first characterize the computation model, agent's benefits, and principal's profits model by considering collaborative training features and user heterogeneity. Furthermore, through the asymmetry of information, we divide the edge intelligent environment into complete information and incomplete information scenarios, in which the principal only knows the distribution of the agents’ private information rather than the specific information. Subsequently by discussing the above two situations in a static environment, we obtain the basic laws of market operation and reveal the significance of introducing a dynamic environment. Then a two-period contract-based incentive mechanism (MotiLearn) is proposed to overcome the ratchet effects of long-term contracts under dynamic incomplete information environment. Finally, theoretical proof and numerical results show that the proposed incentive mechanism is feasible and can motivate both the principal and agents positively.

Традиционни комуникационни модели и нови технологии в PR комуникацията

The entry of digitization into the life of modern society has led to global changes in the economic and socio-communicative spheres. The new dynamic information environment requires the application of special communication skills, as well as new forms of social interaction between organizations and target audiences to achieve high levels of engagement and effectiveness in communication. For a better understanding of the transformation in communication, it is necessary to trace different communication models that are used for a thorough study of the communication process in the field of public relations. It is also important to outline the influence of digitalization to define an integrated environment that implies a dialogic approach and a new type of management of communication processes. The purpose of the publication is to make a comparison between the classical approaches to the study of mass communication and the implementation of new technologies in PR communication, which requires the application of new methods of research and analysis of digital communications, as well as new ways of reaching target groups in social platforms.

Online paths planning method for unmanned surface vehicles based on rapidly exploring random tree and a cooperative potential field

The unstructured, dynamic marine environmental information and the cooperative obstacle avoidance problem greatly challenge the online path planner for unmanned surface vehicles. Efficiency and optimization are crucial for online path planning schemes. Thus, we proposed an algorithmic combination of the optimal rapidly exploring random tree and artificial potential field methods. First, we built a repulsive potential field by considering the relative velocity and position of the unmanned surface vehicle to obstacles and the international regulations for preventing collisions at sea, wherein we designed a repulsive force calculation method using radar readings to avoid irregular obstacles. Then, we guided the sampling process of rapidly exploring random tree using the potential field to accelerate the convergence rate of rapidly exploring random tree to low-cost obstacle avoidance paths. Finally, we planned for multiple paths based on the leader–follower architecture with the guidance of a cooperative potential field. In the experiments, the proposed method consistently outperformed the benchmark methods. We also verified the effectiveness of the algorithmic modifications by conducting ablation experiments.

A modified Q-learning algorithm for robot path planning in a digital twin assembly system

Product assembly is an important stage in complex product manufacturing. How to intelligently plan the assembly process based on dynamic product and environment information has become a pressing issue that needs to be addressed. For this reason, this research has constructed a digital twin assembly system, including virtual and real interactive feedback, data fusion analysis, and decision-making iterative optimization modules. In the virtual space, a modified Q-learning algorithm is proposed to solve the path planning problem in product assembly. The proposed algorithm speeds up the convergence speed by adding a dynamic reward function, optimizes the initial Q table by introducing knowledge and experience through the case-based reasoning algorithm, and prevents entry into the trapped area through the obstacle avoiding method. Finally, the six-joint robot UR10 is taken as an example to verify the performance of the algorithm in the three-dimensional pathfinding space. The experimental results show that the performance of the modified Q-learning algorithm is significantly better than the original Q-learning algorithm in both convergence efficiency and the optimization effect.

A Highly Efficient Multifunctional Wind Barrier Based on PVDF for Power Generation in the Qinghai–Tibet Railway

The Qinghai–Tibet railway is the highest altitude railway globally, and monitoring and maintenance are essential parameters to obtain dynamic structural and environmental information in dangerous areas. Wind barriers are usually used to protect the safe operation of trains at high wind speeds. The power supply for monitoring gadgets is an urgent problem for high hill areas. This article presents a novel and renewable wind barrier for harvesting wind energy. The system architecture includes four main components: 1) the crosswind input module, 2) the vortex shedding occurrence module, 3) the power generation module, and 4) the energy storage and wind load reduction module. Based on the experiments, under the maximum critical wind speed (17 m s−1), the 13.607 V open‐circuit voltage is recorded in a single wind barrier unit with a maximum power of 31.47 μW. The output power is sufficient for charging a 47 μF capacitor to 2.21 V within 50 s. The practicability of the proposed wind barrier for harvesting wind energy is proposed for renewable energy in the Qinghai–Tibet railway.

Exploring the direct and indirect effects of elite influence on public opinion.

Political elites both respond to public opinion and influence it. Elite policy messages can shape individual policy attitudes, but the extent to which they do is difficult to measure in a dynamic information environment. Furthermore, policy messages are not absorbed in isolation, but spread through the social networks in which individuals are embedded, and their effects must be evaluated in light of how they spread across social environments. Using a sample of 358 participants across thirty student organizations at a large Midwestern research university, we experimentally investigate how real social groups consume and share elite information when evaluating a relatively unfamiliar policy area. We find a significant, direct effect of elite policy messages on individuals’ policy attitudes. However, we find no evidence that policy attitudes are impacted indirectly by elite messages filtered through individuals’ social networks. Results illustrate the power of elite influence over public opinion.

Hierarchical speed control for autonomous electric vehicle through deep reinforcement learning and robust control

For the speed control system of autonomous electric vehicle (AEV), challenge happens with how to determine an appropriate driving speed to satisfy the dynamic environment while resisting uncertainty and disturbance. Therefore, this paper proposes a robust optimal speed control approach based on hierarchical architecture for AEV through combining deep reinforcement learning (DRL) and robust control. In decision-making layer, a deep maximum entropy proximal policy optimization (DMEPPO) algorithm is presented to obtain an optimal speed via dynamic environment information, heuristic target entropy and adaptive entropy constraint. In motion control layer, to track the learned optimal speed while resisting uncertainty and disturbance, a robust speed controller is designed by the linear matrix inequality (LMI). Finally, simulation experiment results show that the proposed robust optimal speed control scheme based on hierarchical architecture for AEV is feasible and effective.

A bibliometric review on institutional investor: current status, development and future directions

PurposeIn order to further understand the research status and prospect, the purpose of this paper is to adopt a novel method in the research field of institutional investor to depict the knowledge structure and the evolution path over the past three decades.Design/methodology/approachBased on the 4,194 records retrieved from Web of Science, Citespace combined with VOSviewer are employed to perform visualized analysis.FindingsThe results reveal that the number of published articles of research on institutional investor has an exponential growth. Although the United States is the most significant contributor with more publications compared with other countries, Malaysia and Nigeria show higher centrality in the research network worldwide. Furthermore, “shareholder activism”, “corporate governance”, “global convergence”, “corporate reporting regulation” and “individual investor” are the largest five knowledge clusters. “Media coverage”, “corporate social responsibility” and “stock price crash risk” are the latest three knowledge clusters. Moreover, “governance worldwide”, “institutional character”, “dynamic information environment”, “investment patterns” and “sustainable development” are the potential extended research fields in the future.Originality/valueThis research helps the scholars and participants to capture the knowledge structure of research on institutional investors and to develop a reference to future opportunities.