AI agents -- systems that can independently take actions to pursue complex goals with only limited human oversight -- have entered the mainstream. These systems are now being widely used to produce software, conduct business activities, and automate everyday personal tasks. While AI agents implicate many areas of law, ranging from agency law and contracts to tort liability and labor law, they present particularly pressing questions for the most globally consequential AI regulation: the European Union's AI Act. Promulgated prior to the development and widespread use of AI agents, the EU AI Act faces significant obstacles in confronting the governance challenges arising from this transformative technology, such as performance failures in autonomous task execution, the risk of misuse of agents by malicious actors, and unequal access to the economic opportunities afforded by AI agents. We systematically analyze the EU AI Act's response to these challenges, focusing on both the substantive provisions of the regulation and, crucially, the institutional frameworks that aim to support its implementation. Our analysis of the Act's allocation of monitoring and enforcement responsibilities, reliance on industry self-regulation, and level of government resourcing illustrates how a regulatory framework designed for conventional AI systems can be ill-suited to AI agents. Taken together, our findings suggest that policymakers in the EU and beyond will need to change course, and soon, if they are to effectively govern the next generation of AI technology.
Choosing the type of contract is one of the most important decisions of the host countries for commercial relations with international oil companies. In this regard, in order to attract foreign investment, a contract should be chosen that includes the interests of the parties and with the laws of the host country. Also there is no conflict. In the Kurdistan region of Iraq, despite the same ownership regime as Iran, unlike our country, an attractive contract model, production Sharing, has been used, and it has attracted foreign investment in that region. In this research, the laws and regulations related to the ownership of oil and gas resources in two countries and whether they are in conflict with production sharing contracts or not have been investigated. According to the examination of production Sharing contracts do not conflict with the ownership regime governing oil and gas resources in Iran and Iraq, and regarding the non-transfer of ownership of oil and gas resources in the aforementioned contracts, attention should be paid to three elements related to the exploitation of oil and gas resources, including: mineral rights, extraction rights, and economic rights. In addition, none of the aforementioned rights in the production Sharing Agreement are transferred to the contractor.1. Introduction
Carrying out oil and gas activities, such as upstream activities including: exploration, development and production and Midstream activities including: storage, trading and transportation of oil and gas and downstream activities including: oil refining, marketing and petrochemical production in the oil and gas industry, takes years and requires huge costs and a high level of experience and technology that most governments or national companies do not have. Considering that carrying out the above-mentioned operations and accepting all its risks alone is not possible for many governments with oil and gas resources, and considering that most of the mines are located in developing countries and these countries are in dire need of oil revenue for economic growth and development. In this regard, it is necessary for the government or a national company to establish a business relationship with one or more foreign companies. One of the important issues for establishing the aforementioned relationship is the contract. The main upstream contracts of the oil and gas industry are divided into three categories: concession contracts, sharing contracts, and service contracts. Each of these contracts has a history in Oil Reserves countries, including Iraq and Iran, which are developing countries in need of attracting foreign investment. A: Concession contracts, the use of this contractual method has a history in most Oil Reserves countries in the world. The history of concession contracts in Iraq dates back to a time when the country was part of the Ottoman Empire, and in 1914, the first oil concession was granted by Azam Osmani to the Turkish Oil Company, but World War I prevented its implementation.
In Iran, the first concession contract was concluded in 1858 with the English company "Hotz" for the exploration and exploitation of oil resources. The contract did not achieve the desired result despite drilling several wells on Qeshm Island. (Mirtorabi, 2006, p. 42) B: sharing contracts, the aforementioned contractual system was formed for the first time in Iran and Egypt's oil industry. (Mughraby, 1966, p. 62) The 1957 Petroleum Law introduced and prescribed sharing contracts in Iran for the first time, bypassing the concession contract model, and the first contract concluded based on the aforementioned contractual model in our country was the contract with the Italian company Agip Mineraria in 1957. (Collection of Oil, Gas and Petrochemical Laws and Regulations, 2002, pp 299-327) The use of this contractual model in the Iraq's oil industry was authorized by Law No. 97 of 1967. According to Article 3, Paragraph 2 of the aforementioned law, the National Iraqi Oil Company was authorized to enter into partnerships with international oil companies to extract and exploit fields that were outside the scope of concession contracts. C: Service contracts in Iraq, the first service contract in 1968 was concluded between the National Iraqi Oil Company and Arup Company for the exploration and exploitation of part of the oil resources located in the south of the aforementioned country. Despite oil production in the operational area in 1976, Arup Company withdrew from the aforementioned contract in 1979 in return for receiving compensation. Iran is one of the pioneers of the oil industry in using service contracts in the upstream sector. This contract model was used even before it was prescribed by the legislator. With the explanation is that the first service contract was signed in 1965 between the National Oil Company and the French group Arup. This is while the 1974 Iranian Petroleum Law, which prescribed the aforementioned contract model, was approved eight years later. It is worth mentioning that the contracts concluded must not conflict with the mandatory laws and regulations in the host country. because of the contracts concluded in our country are not attractive to foreign investors, and on the other hand, in the Kurdistan Region of Iraq, due to the same ownership regime as Iran, an attractive contractual model of production sharing is used and this has led to the development of oil and gas resources, the attraction of foreign investment, and the entry of large international companies such as the Norwegian company "DNO" in the tawke oil field, the Genel Energy and China's Sinopec companies in the Taq Taq oil field, and the Dana Gas Company in the Khor Mor gas field. Therefore, the laws and regulations of Iraq and Iran for implementing production participation contracts will be examined by answering the question: What are the legal considerations for implementing production sharing contracts in Iran and Iraq?
2. Methodology
In this research, we tried to collect and gather information from the primary sources (including laws, regulations, contracts, etc.) and secondary sources (such as books and articles, theses, research reports, etc.) in scientific research data centers, official and reputable websites, libraries, and similar resources.
We have chosen Iraq for comparison because, like Iran, it is a developing country with a republican political system and is in need of attracting foreign investment. Considering that Iraq has a similar ownership system to Iran and that an attractive production sharing contract model has been used in the Kurdistan Region of that country, in this regard, first the laws and regulations of Iraq for the implementation of the aforementioned contracts and then the laws and regulations of Iran have been examined.
3. Results and Discussion
Contracts used in the oil and gas industry of countries must not conflict with their laws and regulations. In this regard, one of the important issues for choosing contractual models is the ownership of oil and gas resources. In Iraq, the civil law of the country in question does not directly address the ownership of oil and gas resources, but it states that the owner of the land can fully benefit from it and it also extends to what is above and below the ground. However, Iraq Constitution, according to Article 111, considers oil and gas resources to belong to the public. Therefore, according to the emphasis of the constitution, public ownership governs the oil and gas resources of that country.
Production sharing contracts do not conflict with public ownership of oil and gas resources. Regarding the non-transfer of ownership of oil and gas resources in the aforementioned contracts, attention should be paid to three elements related to the exploitation of oil and gas resources, including mineral rights, extraction rights, and economic rights. As can be seen from the aforementioned names, mineral rights relate to the time when minerals are underground in their original state, extraction rights relate to the rights to extract minerals from the underground, and the rights related to minerals after extraction are economic rights. It should be noted that none of the aforementioned rights in production sharing contracts are transferred to international oil companies, but rather the aforementioned rights belong to the host country, and according to the law, the said country transfers some rights to the national oil company, and the said company, in order to exercise sovereignty, enters into contracts with contractors on behalf of the government to carry out exploration, development, and production operations in the name of the national oil company of the host country.
4. Conclusions and Future Research
Considering that in Iran, contracts concluded in the oil and gas industry have not been very successful in attracting and absorbing foreign investment, the reason for this is that the foreign oil company, which is actually the default loser of the contract, does not have the necessary motivation to implement the win-win and optimal project, and since production sharing contracts in Iraq, despite the dominance of public ownership, are used in the Kurdistan Region of that country and have led to the development of oil and gas resources and also the attraction of investment in that region. Therefore, it is suggested that the aforementioned contracts, which are suitable for attracting and absorbing foreign investment, be used in our country.
Encouraging private sector compliance with ethical data handling practices through incentives and penalties.
Promoting cross-sectoral collaboration between government agencies, tech companies, and advocacy groups to ensure a holistic approach to child data protection.
Conducting periodic assessments and revisions of child data protection policies to keep pace with technological advancements and emerging threats.
Enhancing international cooperation to adopt best practices from global data protection frameworks and ensure seamless integration into Iran’s legal landscape.
By adopting these measures, Iran can create a safer digital landscape for children, ensuring their fundamental right to privacy is protected in the evolving digital age. Addressing legislative gaps and fostering a culture of data privacy awareness will ultimately empower children to navigate the digital world securely and responsibly.
Regulation of industry, trade, and commerce. Occupational law, Islamic law
<i>Background</i>: Customer orders are key in developing logistics processes and in strategic developments of customer orientation. This paper investigates the notion of customer orientation. In the literature, the concept of customer is underdeveloped in that it is seen as a single object rather than how it is enacted in multiple ways. The study examines a customer-oriented production process managed in the supply chain of an automotive manufacturer. <i>Methods</i>: Based on a longitudinal study we explain what constitutes customer knowledge and what processes are involved in constructing this knowledge. <i>Results</i>: The study shows that in a supply chain, multiple versions of customers coexist and overlap that have implications on how multiple-customer orientation is performed and aligned in the supply chain. <i>Conclusions</i>: We find that customer representations proliferate as a result of multiple objectives; we demonstrate what objects and assemblages bring particular customer representation to life and show that these are used to shape customer orientation.
Transportation and communication, Management. Industrial management
Samia Haman, Anass Ben Abdelouahab, Younes El Bouzekri El Idrissi
et al.
<i>Background:</i> The rapid digital transformation driven by Industry 4.0 technologies is reshaping manufacturing supply chains, yet comprehensive insights into how these technologies are integrated remain limited. <i>Methods:</i> This study addresses this research gap by conducting a systematic bibliometric analysis and literature review of integrating Industry 4.0 technologies in the manufacturing supply chain. We used different scientific databases, Scopus and Web of Science, to elaborate this study. <i>Results</i>: Using advanced bibliometric methods, this study examines the evolution of academic discourse, identifies key themes, and maps the intellectual structure of this transformative research field. By leveraging bibliometric tools, the study names the most prolific authors, countries, and journals contributing to this domain. The findings of the first phase reveal the growing focus on topics like supply chain resilience and real-time decision-making, while also finding gaps in the literature related to technology integration. In the second phase, the literature review identified the most used adoption models in empirical studies such as resource-based view, dynamic capabilities view, and technology acceptance model, we also categorized the adoption drivers into technological, organizational, and environmental. <i>Conclusions:</i> This review emphasizes that although research on Industry 4.0 has expanded significantly, the majority of studies predominantly concentrate on technology adoption and quantitative analysis, with little examination of integration, contextual factors, and longitudinal effects.
Transportation and communication, Management. Industrial management
Artificial intelligence (AI) has become integral to various aspects of our lives. While initial advertisements highlighted its numerous benefits, concerns regarding its potential dangers have gained prominence in fields such as social, economic, and legal, etc. The rapid development of AI is to the extent that its nature has been distinguished from mere tools and goods, so the application of object or product liability laws will ignore the unique and real nature of AI. As a means to prevent the associated risks, some researchers have proposed granting legal personality to AI. They draw similarities between AI and companies, suggesting that AI, like corporations, could lead to certain benefits, such as limited liability for its stakeholders. Also, to solve the problem of assets and properties of these technologies, they suggest the use of insurance or the requirement to have minimum assets for AI. Regarding the possible abuse of the beneficiaries, their suggestion is to ignore the legal personality and apply the civil or criminal liability of the beneficiaries. In this way, the incentives for innovation will be greatly increased and the losses of the victims will be compensated in a better way. In this article, we aim to explore and analyze the concept of AI as a legal entity, shedding light on the uncertainties, questions, advantages, and disadvantages surrounding the potential granting of legal personality to AI.1. Introduction
Currently, AI plays an active role in all aspects of human life. Whether functioning as standalone software or integrated with hardware, AI has simplified and improved the way we live. Although not a new technology, recent advancements have captured significant attention. In fact, AI's capabilities are progressing much faster than anticipated, leading to unique traits such as independent decision-making and autonomy. As a result, many now compare AI to humans. This evolution was exemplified in 2017 when Saudi Arabia granted citizenship to a robot named Sophia. While largely symbolic, this event underscored the remarkable progress of AI technology.
Around the same time, the European Union proposed granting AI a legal status termed "electronic personality." Although this proposal faced strong opposition from researchers and experts, it moved the idea of granting legal personality to AI beyond the academic sphere into practical consideration. The basis for this proposal is that current rules regarding accountability and liability for the actions of artificial intelligence are inefficient and inadequate. Present regulations treat these systems merely as tools, ignoring their unique characteristics and potential for independent decision-making.
Proponents argue that granting legal personality to AI could facilitate innovation and investment in the field. The significant differences across various judicial systems, along with the international activities of AI systems, create considerable ambiguity regarding liability for their actions. This uncertainty breeds fear and concern among scientists, investors, and manufacturers. By granting legal personality to AI, stakeholders in AI production and development—including innovators and developers—would have a clearer understanding of the risks and legal consequences of their activities. This increased clarity could, in the long run, attract more investment and resources into the AI industry.
While the European Union's proposal was not explicitly rejected, the EU is currently exploring other methods to address these challenges. However, the discussion this proposal generated has led many researchers to advocate for granting AI legal personality. Given that corporations are considered one of the primary types of legal entities, they have been proposed as a model for bestowing legal personality on AI.
2. Methodology
This research aims to gather information from various sources, including books, articles, theses, research reports, and related materials found in scientific and research data centers, official websites, and libraries. The authors examine two key questions: first, whether there is a scientific basis for granting legal personality to AI; and second, if such a basis exists and legal personality is granted, whether AI has the necessary capabilities to function as a legal person. The analysis of granting legal personality is organized into three sections. The first section explores the nature of artificial intelligence. The second section examines the rationale for granting legal personality, while the third section discusses the advantages and challenges of doing so for AI. This research sought to gather information from various sources, including books, articles, theses, research reports, and similar materials found in scientific and research data centers, official and reputable websites, and libraries. The authors' perspective investigates two key questions: firstly, whether a scientific basis exists for granting legal personality to AI; and secondly, assuming a scientific basis exists and legal personality is granted, whether AI possesses the necessary efficacy to function as a legal person. Therefore, the examination of granting legal personality is organized into three sections. The first section explores the nature of artificial intelligence. The second section examines the basis for granting legal personality. Finally, the third section addresses the advantages and challenges of granting legal personality to AI.
3. Results and Discussion
Proponents of granting legal personality to AI argue that treating AI solely as a product fails to ensure continuous innovation and protect the rights of victims. They define legal personality as a legal construct designed to benefit natural persons, allowing legal systems to allocate specific rights and obligations to this entity as deemed appropriate. Essentially, they view legal personality as a reflection of the legislator's will.
However, bestowing legal personality upon a non-human entity necessitates specific justifications or purposes; the mere appeal or simplification of complex issues is not sufficient. For AI to be considered for legal personality, it must possess certain characteristics, with one of the most critical being the ability to independently own property. The notion of AI holding assets raises several challenges. To address this, proponents suggest implementing insurance or mandating that AI possess a minimum amount of assets. This would require producers and stakeholders to pay insurance premiums or maintain those assets, ensuring that victims receive full compensation without having to deal with a natural person who may lack sufficient resources.
With this approach, AI could be held directly accountable for damages, similar to corporations, while stakeholders (including producers, developers, owners, and operators) would enjoy limited liability akin to corporate directors. Compensation would be confined to the assets of the company, shielding the individuals who created the AI from personal liability for its actions.
Nonetheless, this strategy faces several criticisms. A primary concern is who would be responsible for paying for the insurance or maintaining the minimum assets—the producer or the end-user? Additionally, given the potentially high costs involved, comprehensive insurance coverage may be impractical. There is also the risk that individuals might use AI as a means to evade responsibility.
To counter the potential misuse of legal personality as a shield against liability, proponents suggest applying the "piercing the corporate veil" principle. This principle allows for the dismissal of the legal personality of the AI, holding stakeholders civilly or criminally liable. However, legal systems typically prioritize the social benefits of limited liability and are reluctant to disregard legal personality to expose the identities of its managers. Even if legal systems permitted full application of this principle to AI, advanced, independent AI may not require human oversight or control post-deployment, complicating the identification of responsible parties.
Another challenge arises from the possibility that granting legal personality could discourage producers and users from monitoring their products, as they might feel less accountable for the system's performance. Finally, considering the transnational nature of AI, enforcing legal personality laws across different jurisdictions introduces numerous ambiguities. Granting legal personality to AI in one country while excluding it in another would effectively invalidate that legal personality.
4. Conclusions and Future Research
Granting legal personhood to AI must be grounded in sound reasoning, as it raises both practical and economic considerations. Legal personhood involves not only rights but also obligations. Any AI recognized as a legal person would engage in complex legal relationships that could lead to damages or liabilities. Without genuine accountability, however, such obligations would be meaningless. Additionally, some rights do not require legal personhood; for example, temples in history and animals in current legal systems have enjoyed certain protections without such a designation. Thus, using companies as a model for AI personhood may not be ideal, as AI lacks the organizational structure, decision-making processes, and compliance capacities of corporations.
While granting AI legal personhood could theoretically foster innovation, it also introduces significant risks. AI’s “quasi-shareholders” might lack incentives for safe and cautious product development. Addressing AI-related liabilities requires an approach that aligns with the legal principles of compensation and deterrence. Lawmakers must carefully consider environmental, infrastructural, and technological factors before formally recognizing AI’s legal status. Ultimately, legal systems should first evaluate how existing liability frameworks, preventive measures, and adjustments in product classifications can effectively manage AI’s risks before venturing into the concept of legal personhood.
Therefore, granting personhood to AI could be more harmful than beneficial. In terms of responsibility, it provides no advantage, as stakeholders remain accountable for AI's actions. It's crucial to maintain legal stability; algorithms cannot be recognized as legal entities in one jurisdiction while being ignored in others. Until a comprehensive international legal framework is established, managing AI risks should focus on leveraging existing liability laws, refining product classifications, and implementing robust preventive measures and controls, such as those proposed in the recent European Union AI Act
Regulation of industry, trade, and commerce. Occupational law, Islamic law
<i>Background</i>: The rapid rise of e-commerce has intensified last-mile logistics challenges, fueling the need for sustainable, efficient solutions. Parcel locker crowdshipping systems, integrated with public transport networks, show promise in reducing congestion, emissions, and delivery costs. However, operational and physical constraints (e.g., crowded stations) and liability complexities remain significant barriers to broad adoption. This study investigates the demographic and operational factors that influence the adoption and scalability of these systems. <i>Methods</i>: A mixed-methods design was employed, incorporating survey data from 368 participants alongside insights from 20 semi-structured interviews. Quantitative analysis identified demographic trends and operational preferences, while thematic analysis offered in-depth contextual understanding. <i>Results</i>: Younger adults (18–34), particularly gig-experienced males, emerged as the most engaged demographic. Females and older individuals showed meaningful potential if safety and flexibility concerns were addressed. System efficiency depended on locating parcel lockers within 1 km of major origins and destinations, focusing on moderate parcel weights (3–5 kg), and offering incentives for minor route deviations. Interviews emphasized ensuring that lockers avoid station congestion, clearly defining insurance/liability protocols, and allowing task refusals during peak passenger hours. <i>Conclusions</i>: By leveraging public transport infrastructure, parcel locker crowdshipping requires robust policy frameworks, strategic station-space allocation, and transparent incentives to enhance feasibility.
Transportation and communication, Management. Industrial management
Claudemir Leif Tramarico, Aneirson Francisco Da Silva, José Eduardo Holler Branco
<i>Background:</i> Effective decision-making in supply chain contexts requires understanding how criteria interact to shape rational and transparent decision structures. This study investigates how behavioral aspects influence the structuring of decision-making logic and the interdependencies between key criteria in supply chain contexts. <i>Methods:</i> Using Fuzzy DEMATEL, the research models the interactions between five core criteria —classification, definition, specification, decision, and action feedback—based on inputs from experienced professionals in a global chemical company. The approach enables mapping of causal influences while accounting for subjectivity and uncertainty in expert judgments. <i>Results:</i> The analysis identified specification, definition, and action feedback as causal criteria, with classification and decision being primarily influenced by them. The modeling process supported clearer prioritization and revealed how expert-based interactions can reduce decision biases. <i>Conclusions:</i> This study demonstrates how structuring decision-making logic through causal modeling enhances clarity and reduces subjectivity. The findings contribute to the development of decision support tools applicable across strategic supply chain contexts, offering practical implications for professionals seeking to improve decision transparency and effectiveness.
Transportation and communication, Management. Industrial management
Large language models (LLMs) demand substantial computational and memory resources, creating deployment challenges. Quantization-aware training (QAT) addresses these challenges by reducing model precision while maintaining performance. However, the scaling behavior of QAT, especially at 4-bit precision (W4A4), is not well understood. Existing QAT scaling laws often ignore key factors such as the number of training tokens and quantization granularity, which limits their applicability. This paper proposes a unified scaling law for QAT that models quantization error as a function of model size, training data volume, and quantization group size. Through 268 QAT experiments, we show that quantization error decreases as model size increases, but rises with more training tokens and coarser quantization granularity. To identify the sources of W4A4 quantization error, we decompose it into weight and activation components. Both components follow the overall trend of W4A4 quantization error, but with different sensitivities. Specifically, weight quantization error increases more rapidly with more training tokens. Further analysis shows that the activation quantization error in the FC2 layer, caused by outliers, is the primary bottleneck of W4A4 QAT quantization error. By applying mixed-precision quantization to address this bottleneck, we demonstrate that weight and activation quantization errors can converge to similar levels. Additionally, with more training data, weight quantization error eventually exceeds activation quantization error, suggesting that reducing weight quantization error is also important in such scenarios. These findings offer key insights for improving QAT research and development.
Conformal prediction is a popular framework of uncertainty quantification that constructs prediction sets with coverage guarantees. To uphold the exchangeability assumption, many conformal prediction methods necessitate an additional holdout set for parameter tuning. Yet, the impact of violating this principle on coverage remains underexplored, making it ambiguous in practical applications. In this work, we empirically find that the tuning bias - the coverage gap introduced by leveraging the same dataset for tuning and calibration, is negligible for simple parameter tuning in many conformal prediction methods. In particular, we observe the scaling law of the tuning bias: this bias increases with parameter space complexity and decreases with calibration set size. Formally, we establish a theoretical framework to quantify the tuning bias and provide rigorous proof for the scaling law of the tuning bias by deriving its upper bound. In the end, we discuss how to reduce the tuning bias, guided by the theories we developed.
Joaquin Gonzalez, Liliana Avelar Sosa, Gabriel Bravo
et al.
<i>Background</i>: Efficient inventory management is critical for sustainability in supply chains. However, maintaining adequate inventory levels becomes challenging in the face of unpredictable demand patterns. Furthermore, the need to disseminate demand-related information throughout a company often relies on cloud services. However, this method sometimes encounters issues such as limited bandwidth and increased latency. <i>Methods</i>: To address these challenges, our study introduces a system that incorporates a machine learning algorithm to address inventory-related uncertainties arising from demand fluctuations. Our approach involves the use of an attention mechanism for accurate demand prediction. We combine it with the Newsvendor model to determine optimal inventory levels. The system is integrated with fog computing to facilitate the rapid dissemination of information throughout the company. <i>Results</i>: In experiments, we compare the proposed system with the conventional demand estimation approach based on historical data and observe that the proposed system consistently outperformed the conventional approach. <i>Conclusions</i>: This research introduces an inventory management system based on a novel deep learning architecture that integrates the attention mechanism with cloud computing to address the Newsvendor problem. Experiments demonstrate the better accuracy of this system in comparison to existing methods. More studies should be conducted to explore its applicability to other demand modeling scenarios.
Transportation and communication, Management. Industrial management
Laura Vaccari, Antonio Maria Coruzzolo, Francesco Lolli
et al.
<i>Background:</i> Indoor Positioning Systems (IPS) have gained increasing relevance in logistics, offering solutions for safety enhancement, intralogistics management, and material flow control across various environments such as industrial facilities, offices, hospitals, and supermarkets. This study aims to evaluate IPS technologies’ performance and applicability to guide practitioners in selecting systems suited to specific contexts. <i>Methods:</i> The study systematically reviews key IPS technologies, positioning methods, data types, filtering methods, and hybrid technologies, alongside real-world examples of IPS applications in various testing environments. <i>Results:</i> Our findings reveal that radio-based technologies, such as Radio Frequency Identification (RFID), Ultra-wideband (UWB), Wi-Fi, and Bluetooth (BLE), are the most commonly used, with UWB offering the highest accuracy in industrial settings. Geometric methods, particularly multilateration, proved to be the most effective for positioning and are supported by advanced filtering techniques like the Extended Kalman Filter and machine learning models such as Convolutional Neural Networks. Overall, hybrid approaches that integrate multiple technologies demonstrated enhanced accuracy and reliability, effectively mitigating environmental interferences and signal attenuation. <i>Conclusions:</i> The study provides valuable insights for logistics practitioners, emphasizing the importance of selecting IPS technologies suited to specific operational contexts, where precision and reliability are critical to operational success.
Transportation and communication, Management. Industrial management
Seyed Amin Tabatabaei, Sarah Fancher, Michael Parsons
et al.
We address the task of hierarchical multi-label classification (HMC) of scientific documents at an industrial scale, where hundreds of thousands of documents must be classified across thousands of dynamic labels. The rapid growth of scientific publications necessitates scalable and efficient methods for classification, further complicated by the evolving nature of taxonomies--where new categories are introduced, existing ones are merged, and outdated ones are deprecated. Traditional machine learning approaches, which require costly retraining with each taxonomy update, become impractical due to the high overhead of labelled data collection and model adaptation. Large Language Models (LLMs) have demonstrated great potential in complex tasks such as multi-label classification. However, applying them to large and dynamic taxonomies presents unique challenges as the vast number of labels can exceed LLMs' input limits. In this paper, we present novel methods that combine the strengths of LLMs with dense retrieval techniques to overcome these challenges. Our approach avoids retraining by leveraging zero-shot HMC for real-time label assignment. We evaluate the effectiveness of our methods on SSRN, a large repository of preprints spanning multiple disciplines, and demonstrate significant improvements in both classification accuracy and cost-efficiency. By developing a tailored evaluation framework for dynamic taxonomies and publicly releasing our code, this research provides critical insights into applying LLMs for document classification, where the number of classes corresponds to the number of nodes in a large taxonomy, at an industrial scale.
The exponential growth of open-source package ecosystems, particularly NPM and PyPI, has led to an alarming increase in software supply chain poisoning attacks. Existing static analysis methods struggle with high false positive rates and are easily thwarted by obfuscation and dynamic code execution techniques. While dynamic analysis approaches offer improvements, they often suffer from capturing non-package behaviors and employing simplistic testing strategies that fail to trigger sophisticated malicious behaviors. To address these challenges, we present OSCAR, a robust dynamic code poisoning detection pipeline for NPM and PyPI ecosystems. OSCAR fully executes packages in a sandbox environment, employs fuzz testing on exported functions and classes, and implements aspect-based behavior monitoring with tailored API hook points. We evaluate OSCAR against six existing tools using a comprehensive benchmark dataset of real-world malicious and benign packages. OSCAR achieves an F1 score of 0.95 in NPM and 0.91 in PyPI, confirming that OSCAR is as effective as the current state-of-the-art technologies. Furthermore, for benign packages exhibiting characteristics typical of malicious packages, OSCAR reduces the false positive rate by an average of 32.06% in NPM (from 34.63% to 2.57%) and 39.87% in PyPI (from 41.10% to 1.23%), compared to other tools, significantly reducing the workload of manual reviews in real-world deployments. In cooperation with Ant Group, a leading financial technology company, we have deployed OSCAR on its NPM and PyPI mirrors since January 2023, identifying 10,404 malicious NPM packages and 1,235 malicious PyPI packages over 18 months. This work not only bridges the gap between academic research and industrial application in code poisoning detection but also provides a robust and practical solution that has been thoroughly tested in a real-world industrial setting.
There is strong agreement that generative AI should be regulated, but strong disagreement on how to approach regulation. While some argue that AI regulation should mostly rely on extensions of existing laws, others argue that entirely new laws and regulations are needed to ensure that generative AI benefits society. In this paper, I argue that the debates on generative AI regulation can be informed by the debates and evidence on social media regulation. For example, AI companies have faced allegations of political bias regarding the images and text their models produce, similar to the allegations social media companies have faced regarding content ranking on their platforms. First, I compare and contrast the affordances of generative AI and social media to highlight their similarities and differences. Then, I discuss specific policy recommendations based on the evolution of social media and their regulation. These recommendations include investments in: efforts to counter bias and perceptions thereof (e.g., via transparency, researcher access, oversight boards, democratic input, research studies), specific areas of regulatory concern (e.g., youth wellbeing, election integrity) and trust and safety, computational social science research, and a more global perspective. Applying lessons learnt from social media regulation to generative AI regulation can save effort and time, and prevent avoidable mistakes.
Şemsettin Çiğdem, Ieva Meidute-Kavaliauskiene, Bülent Yıldız
<i>Background:</i> Human–robot collaboration is essential for efficient manufacturing and logistics as robots are increasingly used. Using industrial robots as part of an automation system results in many competitive benefits, including improved quality, efficiency, productivity, and reduced waste and errors. When robots are used in production, human coworkers’ psychological factors can disrupt operations. This study aims to examine the effect of employees’ negative attitudes toward robots on their acceptance of robot technology in manufacturing workplaces. <i>Methods:</i> A survey was conducted with employees in manufacturing companies to collect data on their attitudes towards robots and their willingness to work with them. Data was collected from 499 factory workers in Istanbul using a convenience sampling method, which allowed for the measurement of variables and the analysis of their effects on each other. To analyze the data, structural equation modeling was used. <i>Results:</i> The results indicate that negative attitudes towards robots have a significant negative effect on the acceptance of robot technology in manufacturing workplaces. However, trust in robots was found to be a positive predictor of acceptance. <i>Conclusions:</i> These findings have important implications for manufacturing companies seeking to integrate robot technology into their operations. Addressing employees’ negative attitudes towards robots and building trust in robot technology can increase the acceptance of robots in manufacturing workplaces, leading to improved efficiency and productivity.
Transportation and communication, Management. Industrial management
Ślęzak Marcin, Szczepański Tomasz, Stasiak-Cieślak Beata
et al.
The article concerns the problem of lateral stabilization of a tricycle with variable front wheel track. The vehicle can operate in two modes: with the front wheels folded out and connected. A bicycle designed for people with special needs. The element that requires elaboration is the stabilization while driving with the front wheels unfolded. In this mode, a metastable state is created. When the lateral angle of the slope exceeds the limit value, the lateral force causes the lateral tilt. This phenomenon should be treated as a disadvantage as it makes it difficult to smoothly tilt the bicycle, which can lead to disorientation of the rider and difficulty in maintaining balance. The article presents mathematical simulations that allow for the analysis of factors influencing the discussed phenomenon.
This paper explores the evolution of China's Personal Information Protection Law (PIPL) and situates it within the context of global data protection development. It draws inspiration from the theory of 'Brussels Effect' and provides a critical account of its application in non-Western jurisdictions, taking China as a prime example. Our objective is not to provide a comparative commentary on China's legal development but to illuminate the intricate dynamics between the Chinese law and the EU's GDPR. We argue that the trajectory of China's Personal Information Protection Law calls into question the applicability of the Brussels Effect: while the GDPR's imprint on the PIPL is evident, a deeper analysis unveils China's nuanced, non-linear adoption that diverges from many assumptions of the Brussels Effect and similar theories. The evolution of the GDPR-inspired PIPL is not as a straightforward outcome of the Brussels Effect but as a nuanced, intricate interplay of external influence and domestic dynamics. We introduce a complementary theory of 'gravity assist', which portrays China's strategic instrumentalisation of the GDPR as a template to shape its unique data protection landscape. Our theoretical framework highlights how China navigates through a patchwork of internal considerations, international standards, and strategic choices, ultimately sculpting a data protection regime that has a similar appearance to the GDPR but aligns with its distinct political, cultural and legal landscape. With a detailed historical and policy analysis of the PIPL, coupled with reasonable speculations on its future avenues, our analysis presents a pragmatic, culturally congruent approach to legal development in China. It signals a trajectory that, while potentially converging at a principled level, is likely to diverge significantly in practice [...]
We present a fast, flexible heuristic for setting up warehouse locations for quick commerce businesses, with the goal of serving the largest number of customers under the constraints of delivery radius and maximum daily deliveries per warehouse. Quick commerce or direct-to-customer delivery businesses guarantee delivery within a specified time. Using experiments on various scenarios, we show that the proposed algorithm is flexible enough to handle variations such as non-uniform population distributions, variable travel times, and selection of multiple warehouse locations.
So-called 'fast fashion' consumption, amplified through cost-effective e-commerce, constitutes a major factor negatively impacting climate change. A recently noted strategy to motivate consumers to more sustainable decisions is digital nudging. This paper explores the capability of digital nudging in the context of green fashion e-commerce. To do so, digital default and social norm nudges are tested in an experimental setting of green fashion purchases. An online experiment (n=320) was conducted, simulating an online retail scenario. Results failed to show statistically significant relationships between nudging strategies and purchase decisions. However, explorative analyses show a backfiring effect for the combination of nudges and thus, reveal a hitherto neglected impact of participants' identification on the effectiveness of the digital nudging strategies. Consequently, this study contributes to digital nudging literature and informs practice with new insights on effective choice architectures in e-commerce.