The paper proposes a new approach to evaluating the quality of public transport based on underperformance analysis. This approach aims to identify actual service failures by combining objective data and user perceptions. In this context, the objective of this paper is to offer a more reliable evaluation tool adapted to the contexts of developing countries, where certain service deficiencies tend to be normalized. The proposed approach is applied to a real case study in Constantine public transport; a series of objective and subjective indicators are calculated on the basis of infractions provided by local authorities and user perceptions of the service. This approach thus allows for a better understanding of the sources of quality degradation and supports more effective decisions for improving public transport in order to achieve the overall objectives of sustainable mobility.
<i>Background</i>: The constant growth in demand for sustainable energy products and the development of the circular economy have created a critical need for an efficient supply chain for biomass. However, the inherent challenges of biomass make its harvesting, collection, storage, and transport difficult, impacting logistical efficiency and the viability of bioenergy and bioproduct production. This study analyzes how combining artificial intelligence (AI) with multimodal transport can optimize and improve efficiency, as well as reduce costs, in biomass logistics. <i>Methods</i>: The study uses a tiered research framework that encompasses the physical domain (biomass limitations), the structural domain (mathematical modeling for multimodal transport), the intelligence domain (AI-based decision making), and the strategic approach. <i>Results</i>: The outcomes indicate that while truck transport is ideal for short distances, integrating rail and water transport through AI-driven optimization reduces costs and greenhouse gas emissions for long-distance travel. AI technologies, such as digital twins and machine learning, improve demand forecasting, real-time routing, and cargo consolidation, leading to enhanced prediction accuracy for transport costs. <i>Conclusions</i>: The integration of AI and multimodal networks builds resilient and sustainable biomass supply chains. However, full implementation requires addressing data fragmentation and investing in digital infrastructure to enable seamless coordination between supply chain stakeholders.
Transportation and communication, Management. Industrial management
This paper extends the Acemoglu-Restrepo task exposure framework to address the labor market effects of agentic artificial intelligence systems: autonomous AI agents capable of completing entire occupational workflows rather than discrete tasks. Unlike prior automation technologies that substitute for individual subtasks, agentic AI systems execute end-to-end workflows involving multi-step reasoning, tool invocation, and autonomous decision-making, substantially expanding occupational displacement risk beyond what existing task-level analyses capture. We introduce the Agentic Task Exposure (ATE) score, a composite measure computed algorithmically from O*NET task data using calibrated adoption parameters--not a regression estimate--incorporating AI capability scores, workflow coverage factors, and logistic adoption velocity. Applying the ATE framework across five major US technology regions (Seattle-Tacoma, San Francisco Bay Area, Austin, New York, and Boston) over a 2025-2030 horizon, we find that 93.2% of the 236 analyzed occupations across six information-intensive SOC groups (financial, legal, healthcare, healthcare support, sales, and administrative/clerical) cross the moderate-risk threshold (ATE >= 0.35) in Tier 1 regions by 2030, with credit analysts, judges, and sustainability specialists reaching ATE scores of 0.43-0.47. We simultaneously identify seventeen emerging occupational categories benefiting from reinstatement effects, concentrated in human-AI collaboration, AI governance, and domain-specific AI operations roles. Our findings carry implications for workforce transition policy, regional economic planning, and the temporal dynamics of labor market adjustment
Artificial intelligence is a emerging phenomenon in the field of technology that, according to scientists' predictions, will significantly transform many aspects of life in the near future and bring about important changes in the course of human existence. With the continuous advancement of technology, questions arise regarding its impacts and roles in everyday life and society. This research aims to answer the fundamental question of the conflicts that artificial intelligence faces with human rights and citizenship if it possesses legal personality, using a descriptive-analytical method and collecting the necessary information and resources through a library-based approach. The findings of the research indicate that, according to researchers and scholars in this field, granting legal personality to artificial intelligence is necessary. They believe that by granting personality, AI entities can become parties to contracts, own property, and bear responsibilities. However, some argue that artificial intelligence should not be granted legal personality, as it would create ambiguity in the boundary between humans and non-humans and it undermines human rights and citizenship with its interferences and conflicts. Therefore, they suggest that AI should remain under human control. Furthermore, this research has proposed models for the legal personality of artificial intelligence, and it has identified the challenges and conflicts that AI faces in relation to human rights and citizenship, such as Automatic performance of military weapons, privacy violations, job displacement, social manipulation, fake news, and more. Solutions to address these challenges have also been outlined.
1. Introduction
The issue of the legal status of artificial intelligence has gained attention in today's society. In recent years, AI has made significant advancements as machines that imitate human thinking and perform tasks automatically. Consequently, the legal implications of AI are increasingly unknown. The legal status of AI encompasses rights, responsibilities, and accountability. To ensure the responsible and ethical use of AI, its legal status must first be clarified and addressed. Due to the diverse and constantly changing nature of artificial intelligence technologies, defining and classifying AI systems is challenging. The terms and criteria related to AI have not yet been clearly defined. Therefore, such ambiguity makes it difficult to formulate laws and frameworks for AI.
Determining legal responsibilities for the actions and functions of AI systems is complex. Traditional legal systems assign responsibility based on intention and action. When AI systems become more independent, which individual or entity is responsible for the damages or incidents caused by AI? Thus, when AI systems learn and evolve independently or exhibit behaviors outside of their programming, this issue becomes even more complicated. Ethics and law in artificial intelligence are interconnected. AI systems can impact human lives; therefore, the development and implementation of ethics is essential. The functioning of AI raises concerns regarding fairness, transparency, privacy, and discrimination. Addressing these ethical issues and ensuring that AI systems respect human rights and values necessitates the legal regulation of AI.
Inventions and content generated by AI complicate intellectual property rights. The legal status of AI affects trade secrets, copyright ownership, and patentability. Existing legal frameworks must ensure a balance between promoting innovation and protecting creators and innovators in the field of AI. The rapid growth of artificial intelligence and its unique characteristics necessitate a thorough examination of the legal framework. The complexities and threats posed by AI may fall outside current laws. AI may require new ideas, systems, and legal standards to address its specific features and challenges.
It is essential to consider the legal status of AI at the international level, which requires international cooperation. Data protection and international standards for the development and implementation of AI need coordination within regulatory frameworks across jurisdictions. Conventions, laws, and global frameworks for responsible and ethical AI practices require collaboration. In recent years, there has been significant research focused on AI technology, thus legal discussions must also be addressed. So far, there has been no comprehensive study on the legal status of AI, its interference and conflict with collective human rights and civil rights, the challenges it presents, and potential solutions. Most research conducted has examined its functionality and applications in areas such as insurance, judicial proceedings, healthcare, Islamic studies, and humanities.
Therefore, this study first discusses the differences between artificial intelligence and humans, followed by concepts and definitions of AI to provide a basic understanding for legal professionals (without offering highly specialized definitions). It then addresses the legal status of AI, theories regarding its acceptance or rejection, predictions of personality models for AI, instances where AI may conflict with ethical and human laws, and strategies to prevent such conflicts. The research method used in this article is descriptive-analytical.
2. Methodology
This research has been conducted using a descriptive-analytical method, and the necessary information and resources have been collected through library research.
3. Results and Discussion
In the future, artificial intelligence will advance significantly and surpass its current position, so considering this, it still requires important research. In addition to opinions that believe AI might replace humans by impersonating their identity and personality, necessitating control, some views suggest this technology will lead to significant benefits for humanity.
All these discussions can be resolved by establishing laws regarding the granting of personhood to AI. By granting personhood to AI in the form of a general legal status, AI will be able to have rights and obligations through its actions.
Furthermore, if AI causes harm to another person due to its actions, liability may arise. Thus, due to the granting of personhood to AI, it will be able to participate in a case and even act as a lawyer or representative in a lawsuit. However, it should not be forgotten that no matter how advanced AI becomes, it must adhere to certain limitations in its personhood status, similar to legal entities.
In the context of the legal status of artificial intelligence in Iran, there are currently no definitive regulations. When examining the discussions that have taken place on this subject, some viewpoints consider it appropriate to treat artificial intelligence as an object, grant it legal personality, or recognize it as a non-human, electronic, or artificial person. In fact, all the debates that have occurred in this regard are deemed valuable and important, each providing significant arguments for integrating the concept of artificial intelligence into the legal system.
However, following the European Parliament's proposal to allocate electronic personality and important responsibilities, the dimensions of the discussion have changed. Although many issues still remain to be clarified. It is unclear to what extent and in which matters the European Parliament will approve this report. Similarly, apart from recognizing legal personality, there is no clear information regarding the limitations and exceptions of this personality. Therefore, from the perspective of international law, it is essential to urgently develop legal regulations concerning the legal status of artificial intelligence within the United Nations or the European Union.
In this context, doctrinal discussions and the examination of advantages and disadvantages are the most important data that will facilitate the improvement of legal regulations. In contrast to artificial intelligence, which is becoming systematically more autonomous and intelligent, legal personality must undoubtedly be accepted. However, the concept of such personality should be a unique one for artificial intelligence.
4. Conclusions and Future Research
Intelligent robots can pose challenges for humans and their human rights, including but not limited to the following: violations of privacy, the potential for the automatic operation of dangerous military weapons, social interventions, and the spread of fake and misleading news and information, increased unemployment and reduced job opportunities, gender and racial discrimination, inequality in matters related to criminal justice, healthcare, hiring, and more. Therefore, it is necessary to devise solutions to address these challenges, which may include diversifying data, ensuring operational transparency, establishing legal and ethical frameworks, government commitments to balance the use of artificial intelligence with the fulfillment and enforcement of the right to work, data protection, respect for privacy, reducing discrimination, collaborating with stakeholders, establishing international legal standards, and fostering international cooperation for the safe development of artificial intelligence. Otherwise, it would be an affirmation of human rights violations by artificial intelligence.
Regulation of industry, trade, and commerce. Occupational law, Islamic law
Mona ElSemary, Nada Eman, Dana Corina Deselnicu
et al.
<i>Background</i>: nowadays, traditional delivery options are challenging to the urban last-mile logistics and sustainability goals. The purpose of this study is to investigate the practical factors that drive frequent e-shoppers to actively switch their intention from conventional delivery options to utilizing smart lockers. <i>Methods</i>: the hypothetical framework tested integrating constructs from the Technology Acceptance Model (TAM) and the Theory of Planned Behavior (TPB), and supplementary constructs such as privacy and convenience. Data were collected via a structured online questionnaire from 513 respondents in major Egyptian cities, including Alexandria and Cairo. The framework was tested using Structural Equation Modeling (SEM) via SmartPLS 4.0 software to assess the relationship between constructs and switching intention. <i>Results</i>: the analysis confirms that switching intention to use smart lockers is positively driven by Perceived Usefulness, Perceived Ease of Use, Convenience, Privacy, and Perceived Behavioral Control. Notably, a positive attitude towards smart lockers was found to have a non-significant effect on the intention to switch in the Egyptian context. <i>Conclusions</i>: this research contributes to addressing the gap in the extant literature by focusing on analyzing the unique contextual determinants in the emerging last-mile logistics within a developing market context.
Transportation and communication, Management. Industrial management
We consider the impact of fairness requirements on the social efficiency of truthful mechanisms for trade, focusing on Bayesian bilateral-trade settings. Unlike the full information case in which all gains-from-trade can be realized and equally split between the two parties, in the private information setting, equitability has devastating welfare implications (even if only required to hold ex-ante). We thus search for an alternative fairness notion and suggest requiring the mechanism to be KS-fair: it must ex-ante equalize the fraction of the ideal utilities of the two traders. We show that there is always a KS-fair (simple) truthful mechanism with expected gains-from-trade that are half the optimum, but always ensuring any better fraction is impossible (even when the seller value is zero). We then restrict our attention to trade settings with a zero-value seller and a buyer with value distribution that is Regular or MHR, proving that much better fractions can be obtained under these conditions.
This paper provides an overview and critique of the risk based model of artificial intelligence (AI) governance that has become a popular approach to AI regulation across multiple jurisdictions. The 'AI Policy Landscape in Europe, North America and Australia' section summarises the existing AI policy efforts across these jurisdictions, with a focus of the EU AI Act and the Australian Department of Industry, Science and Regulation's (DISR) safe and responsible AI consultation. The 'Analysis' section of this paper proposes several criticisms of the risk based approach to AI governance, arguing that the construction and calculation of risks that they use reproduces existing inequalities. Drawing on the work of Julia Black, it argues that risk and harm should be distinguished clearly and that the notion of risk is problematic as its inherent normativity reproduces dominant and harmful narratives about whose interests matter, and risk categorizations should be subject to deep scrutiny. This paper concludes with the suggestion that existing risk governance scholarship can provide valuable insights toward the improvement of the risk based AI governance, and that the use of multiple regulatory implements and responsive risk regulation should be considered in the continuing development of the model.
Jon Altonaga Puente, Enrico Mezzetti, Irune Agirre Troncoso
et al.
Multicore timing interference, arising when multiple requests contend for the same shared hardware resources, is a primary concern for timing verification and validation of time-critical applications. Bandwidth control and regulation approaches have been proposed in the literature as an effective method to monitor and limit the impact of timing interference at run time. These approaches seek for fine-grained control of the bandwidth consumption (at the microsecond level) to meet stringent timing requirements on embedded critical systems. Such granularity and configurations, while effective, can become an entry barrier for the application of bandwidth control to a wide class of productized, modular ROS2 applications. This is so because those applications have less stringent timing requirements but would still benefit from bandwidth regulation, though under less restrictive, and therefore more portable, granularity and configurations. In this work, we provide ROSGuard, a highly-portable, modular implementation of a timing interference monitoring and control mechanism that builds on the abstractions available on top of a generic and portable Linux-based software stack with the Robotic Operating System 2 (ROS2) layer, a widespreadedly adopted middleware for a wide class of industrial applications, far beyond the robotic domain. We deploy ROSGuard on an NVIDIA AGX Orin platform as a representative target for functionally rich distributed AI-based applications and a set of synthetic and real-world benchmarks. We apply an effective bandwidth regulation scheme on ROS2-based applications and achieve comparable effectiveness to specialized, finer-grained state-of-the-art solutions.
This study offers a comprehensive analysis of the evolving landscape of International Trade Law in response to dynamic global forces. Tracing the historical foundations through mercantilism to the contemporary era, the research explores key developments such as the creation of the World Trade Organization, the rise of digitalization, and the integration of environmental and social considerations. Recent legal shifts are examined in terms of the broadening scope, enhanced structure, and augmented complexity of International Trade Law. The study highlights the transformative impact of the COVID-19 pandemic and provides valuable insights for businesses, policymakers, and researchers navigating the complexities of an interconnected and dynamic global economy. As International Trade Law adapts to emerging challenges, the study underscores its pivotal role in fostering fairness, justice, and global economic cooperation.
Benjamin Mosses Sakita, Berit Irene Helgheim, Svein Bråthen
<i>Background</i>: Scholarly literature indicates a slow pace at which maritime ports fully embrace digital transformation (DT). The reasons to this are largely anecdotal and lack solid empirical grounding. This inhibits an overall understanding of DT’s tenets and the development of evidence-based policies and targeted actions. <i>Methods</i>: This study deployed a qualitative case study strategy to unpack the challenges of undertaking DT through the lens of principal-agent theory (PAT). <i>Results</i>: Analysis of data collected through 13 semi-structured interviews from a port’s value chain stakeholders revealed five thematic challenges that contradict successful implementation of DT. These included interagency constraints and system ownership tussles; system sabotage and prevalent corruption; prevalent human agency in port operations; cultural constraints; and political influence on port governance. <i>Conclusions</i>: To address these challenges, the study proposes a four-stage empirically grounded DT strategy framework that guides both practitioners and policymakers through DT endeavors. The framework includes: (1) the port’s value chain mapping, (2) stakeholder engagement, (3) resource mobilization, and (4) effective monitoring. For scholars, we provide an avenue for testing statistical significance of association and causality among the identified challenges.
Transportation and communication, Management. Industrial management
<i>Background:</i> Despite the fact that some results can be found for the logistics simulation in the literature, there is a lack of an experiment based on inner company data. Therefore, the study aimed to determine the potential need for this kind of solution by Zala County enterprises. <i>Methods:</i> As a first step, the paper presents the existing literature with the help of a literature review. Afterwards, questionnaire sampling was conducted among local enterprises. The paper applies several statistical methods (e.g., descriptive statistics, SPSS, exploratory factor analysis, and structural equation modeling) to the sample for the validation of the reorganization demand. <i>Results:</i> The study obtained a total of 147 complete responses from the 1022 invitations. An overwhelming majority of the respondents indicated regular and significant delays in their logistics processes, as well as the need for a new simulation method. Based on the SEM model, it has been observed that IT solutions are being utilized in an inefficient manner, resulting in logistical system issues and operational damages. <i>Conclusion</i>: The paper successfully identified a research gap, the research of which can have not only theoretical but also many practical benefits. Hopefully, the results will generate other academic research in this field.
Transportation and communication, Management. Industrial management
David Wichner, Jeffrey Wishart, Jason Sergent
et al.
Safety Management Systems (SMSs) have been used in many safety-critical industries and are now being developed and deployed in the automated driving system (ADS)-equipped vehicle (AV) sector. Industries with decades of SMS deployment have established frameworks tailored to their specific context. Several frameworks for an AV industry SMS have been proposed or are currently under development. These frameworks borrow heavily from the aviation industry although the AV and aviation industries differ in many significant ways. In this context, there is a need to review the approach to develop an SMS that is tailored to the AV industry, building on generalized lessons learned from other safety-sensitive industries. A harmonized AV-industry SMS framework would establish a single set of SMS practices to address management of broad safety risks in an integrated manner and advance the establishment of a more mature regulatory framework. This paper outlines a proposed SMS framework for the AV industry based on robust taxonomy development and validation criteria and provides rationale for such an approach. Keywords: Safety Management System (SMS), Automated Driving System (ADS), ADS-Equipped Vehicle, Autonomous Vehicles (AV)
Fernando Miguelez, Josu Doncel, Maria Dolores Ugarte
Industrial processes generate a massive amount of monitoring data that can be exploited to uncover hidden time losses in the system. This can be used to enhance the accuracy of maintenance policies and increase the effectiveness of the equipment. In this work, we propose a method for one-step probabilistic multivariate forecasting of time variables involved in a production process. The method is based on an Input-Output Hidden Markov Model (IO-HMM), in which the parameters of interest are the state transition probabilities and the parameters of the observations' joint density. The ultimate goal of the method is to predict operational process times in the near future, which enables the identification of hidden losses and the location of improvement areas in the process. The input stream in the IO-HMM model includes past values of the response variables and other process features, such as calendar variables, that can have an impact on the model's parameters. The discrete part of the IO-HMM models the operational mode of the process. The state transition probabilities are supposed to change over time and are updated using Bayesian principles. The continuous part of the IO-HMM models the joint density of the response variables. The estimate of the continuous model parameters is recursively computed through an adaptive algorithm that also admits a Bayesian interpretation. The adaptive algorithm allows for efficient updating of the current parameter estimates as soon as new information is available. We evaluate the method's performance using a real data set obtained from a company in a particular sector, and the results are compared with a collection of benchmark models.
The subject of this paper is the AvatarTraffic simulator – a computer system capable of modelling in real time environments such as subway stations or airport halls populated with tens or hundreds of moving figures, which, in addition to pedestrian traffic typical for this type of objects, can perform predefined sequences of events and actions formulated as a simulation scenario. Thanks to the integration with a real monitoring system, the simulator, in addition to providing data streams (including video) generated by the virtual scene, is also able to dynamically respond to actions taken by the system’s staff. Using the Unity simulation engine as the implementation platform, a number of practical problems had to be solved during the development, two of which are the subject of this article: a) supervising and correcting the work of AI algorithms used in Unity to simulate the pedestrian movement of avatars, and b) a textual description of the scenario of events taking place on the stage in a way editable for experts planning tests of the monitoring system. Some more challenging cases of people movement are discussed (including creating queues and passing through doors) and the paper presents original algorithms correcting the work of the Unity’s built-in methods in the situations when the coordinated behaviour of people groups is required. Because of the specifics of the simulator environment the scenario needed to be expressed in a JSON text file, and the article presents the implemented mechanisms of its compilation directly to the C# runtime environment and discusses the original command language which was created to model sequences of events and actions making up the scenario.
Bolimera Anudeepsekhar, Muthalagu Raja, Kalaichelvi V.
et al.
Ego lane detection is one of the key techniques in Ego Lane Analysis System (ELAS) implemented in smart autonomous driving cars for lane detection in roads. This technique has been extensively studied in recent years because of its accurate and robust detection of shape and location of lanes. The conventional methods are less robust and computationally expensive since they have several challenges in localization of lanes due to presence of occlusions on roads. So to avoid these issues, this paper uses a novel 2-stage lane detection method using deep convolutional neural network to detect the lanes and its key-points by optimally fit a curve to the lane to compensate on above mentioned issues. The proposed methodology for lane detection predicts the key-points accurately and it robust under various weather conditions and highway driving scenarios. In terms of performance, this technique is fast and robust with low computational cost and has high performance when deployed on autonomous vehicle-based systems.
<i>Background:</i> Nowadays, as a result of globalization, markets are more competitive, and customers are more demanding. To respond to these challenges, organizations must develop mechanisms for continuous improvement in order to eliminate waste and increase the efficiency and effectiveness of processes. Thus, the present study carried out at an industrial unit responsible for the customization of cork stoppers for wines had as its main objectives to identify and eliminate or at least reduce waste; improve production and internal logistics flows; balance workloads; improve productivity; reduce lead time; motivate employees and promote the spirit of continuous improvement. <i>Methods:</i> The action-research methodology was used, whereby several cycles of data recovery and analysis, identification and implementation of opportunities for improvement, assessment and standardization were carried out. Therefore, the Total Flow Management (TFM) model was implemented, and several methods and tools were used, such as Value Stream Mapping (VSM), work measurement and 5S’s. <i>Results:</i> Several wastes and overloads were identified, and some actions were implemented, such as workload balancing, layout changes, implementation of visual management and supermarkets. That said, it was possible to reduce lead time by 4 days, improve productivity from 26.63 ML (a thousand cork stoppers)/h to 35.75 ML/h, and promote flexibility. In addition, employees were motivated, and a culture of continuous improvement was fostered. <i>Conclusions:</i> This project demonstrated that it is possible to implement improvement actions, with good results, without high investments, as well as motivating employees and taking advantage of their best capabilities. Additionally, it was demonstrated that the use of TFM can be very useful in continuous improvement, with evident improvements in production and internal logistics flows. So, this project demonstrated the practical implementation of TFM regarding basic reliability, production and internal logistics flow, and the simultaneous use of several methods and tools to implement continuous improvement. Thus, significant improvements were possible on the factory floor, as well as improving employee motivation their personal development and encouraging the focus on continuous improvement. Therefore, it responds to the gap identified in the literature.
Transportation and communication, Management. Industrial management
Ebenezer Laryea, Amin Hosseinian-Far, Simon Derrick
<i>Background</i>: Airfreight transport refers to the shipment of goods by air from one location to another and is often perceived as a contributor to global carbon emissions. The environmental impacts associated with airfreight are of notable and genuine concern. Such concerns have often led to calls for measures to ban or limit air freight as a mode of transportation for goods. Whilst the majority of these calls are perceived to be well placed, it is nevertheless essential to acknowledge the climate justice implications associated with such measures, particularly in the context of perishable products like fresh produce. <i>Methods</i>: The aim of this study is to thoroughly examine the socioeconomic implications of banning air-freighted fresh produce and to recommend practices that can minimize the environmental impacts. Utilizing Blue Skies Holdings Ltd., Pitsford, UK as a case study, this paper undertakes a comprehensive analysis of the potential climate justice ramifications associated with the prohibition of air-freighted fresh produce. <i>Results</i>: The analysis highlights the intricate interplay between the environmental and socioeconomic dimensions of the issue. By investigating the carbon emissions attributed to aviation and air logistics in particular and meticulously scrutinizing the possible consequences of an airfreight ban in relation to vulnerable communities within developing economies that are heavily reliant on fresh produce exports, the study contributes insights to guide policy discourse and the decision-making processes within commercial entities with respect to their carbon emissions reduction strategies. <i>Conclusions</i>: Accordingly, this study provides a number of recommendations for various actors, particularly commercial stakeholders, who deal with air-freighted fresh produce.
Transportation and communication, Management. Industrial management
Personalizing user experience with high-quality recommendations based on user activity is vital for e-commerce platforms. This is particularly important in scenarios where the user's intent is not explicit, such as on the homepage. Recently, personalized embedding-based systems have significantly improved the quality of recommendations and search in the e-commerce domain. However, most of these works focus on enhancing the retrieval stage. In this paper, we demonstrate that features produced by retrieval-focused deep learning models are sub-optimal for ranking stage in e-commerce recommendations. To address this issue, we propose a two-stage training process that fine-tunes two-tower models to achieve optimal ranking performance. We provide a detailed description of our transformer-based two-tower model architecture, which is specifically designed for personalization in e-commerce. Additionally, we introduce a novel technique for debiasing context in offline models and report significant improvements in ranking performance when using web-search queries for e-commerce recommendations. Our model has been successfully deployed at Yandex, serves millions of users daily, and has delivered strong performance in online A/B testing.
This article presents a novel solution for reconfigurable intelligent surfaces (RISs) based on cascaded channel decoupling. The proposed mechanism simplifies the RIS regulation matrix, by decomposing the electromagnetic wave regulation process into two sub-processes: virtual receiving response and virtual regular transmission, which leads to the decoupling of the RIS cascaded channel. This article further discusses the concrete implementation of the proposed channel decoupling mechanism in two scenarios of single-user access and multi-user access, and gives the corresponding detailed scheme. The numerical simulation results demonstrate that the proposed channel decoupling scheme is a low-complexity and effective solution for resolving the RIS regulation matrix.
Industry 4.0 operates based on IoT devices, sensors, and actuators, transforming the use of computing resources and software solutions in diverse sectors. Various Industry 4.0 latency-sensitive applications function based on machine learning to process sensor data for automation and other industrial activities. Sending sensor data to cloud systems is time consuming and detrimental to the latency constraints of the applications, thus, fog computing is often deployed. Executing these applications across heterogeneous fog systems demonstrates stochastic execution time behavior that affects the task completion time. We investigate and model various Industry 4.0 ML-based applications' stochastic executions and analyze them. Industries like oil and gas are prone to disasters requiring coordination of various latency-sensitive activities. Hence, fog computing resources can get oversubscribed due to the surge in the computing demands during a disaster. We propose federating nearby fog computing systems and forming a fog federation to make remote Industry 4.0 sites resilient against the surge in computing demands. We propose a statistical resource allocation method across fog federation for latency-sensitive tasks. Many of the modern Industry 4.0 applications operate based on a workflow of micro-services that are used alone within an industrial site. As such, industry 4.0 solutions need to be aware of applications' architecture, particularly monolithic vs. micro-service. Therefore, we propose a probability-based resource allocation method that can partition micro-service workflows across fog federation to meet their latency constraints. Another concern in Industry 4.0 is the data privacy of the federated fog. As such, we propose a solution based on federated learning to train industrial ML applications across federated fog systems without compromising the data confidentiality.