The development of large-scale foundation models, particularly Large Language Models (LLMs), is constrained by significant computational and memory bottlenecks. These challenges elevate throughput optimization from a mere engineering task to a critical strategic lever, directly influencing training time, operational cost, and the feasible scale of next-generation models. This paper synthesizes evidence from recent academic and industry innovations to analyze key advancements in training efficiency. We examine architectural solutions to dataloader bottlenecks, such as the OVERLORD framework, which has demonstrated a 4.5% improvement in end-to-end training throughput. We investigate memory optimization techniques designed to overcome the GPU memory wall, including CPU offloading strategies like DeepSpeed's ZeRO-Offload, which enable the training of models far exceeding single-accelerator capacity. Furthermore, we explore the growing importance of compiler-centric optimizations, exemplified by Triton-distributed, which enables the joint optimization of computation, memory, and communication for substantial performance gains. The analysis is contextualized by advanced profiling tools and hardware characterization studies that identify and mitigate previously overlooked overheads like Dynamic Voltage and Frequency Scaling (DVFS). Findings indicate that a holistic, system-level approach, integrating innovations across data pipelines, memory management, network fabrics, and compiler technologies, is essential for accelerating AI development, managing costs, and pushing the boundaries of model scale.
Teaching integrated with environmental issues has strengthened the cross-cutting nature of different subjects in basic education. Thus, this study aimed to report the results of implementing a teaching sequence focused on teaching/learning electrodynamics, based on the debate about sustainable electricity generation. The teaching sequence involved discussions on the environment, electricity generation, and sustainable development, culminating in the construction of a model composed of lamps, LEDs, resistors, motors, and actuators. The proposal was implemented in classes of high school students at a full-time state school. The proposal proved effective, promoting students' understanding of the content and reflection on the topic.
Large Language Models (LLMs) are catalyzing a paradigm shift in scientific discovery, evolving from task-specific automation tools into increasingly autonomous agents and fundamentally redefining research processes and human-AI collaboration. This survey systematically charts this burgeoning field, placing a central focus on the changing roles and escalating capabilities of LLMs in science. Through the lens of the scientific method, we introduce a foundational three-level taxonomy-Tool, Analyst, and Scientist-to delineate their escalating autonomy and evolving responsibilities within the research lifecycle. We further identify pivotal challenges and future research trajectories such as robotic automation, self-improvement, and ethical governance. Overall, this survey provides a conceptual architecture and strategic foresight to navigate and shape the future of AI-driven scientific discovery, fostering both rapid innovation and responsible advancement. Github Repository: https://github.com/HKUST-KnowComp/Awesome-LLM-Scientific-Discovery.
This paper explores the critical transition from Generative Artificial Intelligence (GenAI) to Innovative Artificial Intelligence (InAI). While recent advancements in GenAI have enabled systems to produce high-quality content across various domains, these models often lack the capacity for true innovation. In this context, innovation is defined as the ability to generate novel and useful outputs that go beyond mere replication of learned data. The paper examines this shift and proposes a roadmap for developing AI systems that can generate content and engage in autonomous problem-solving and creative ideation. The work provides both theoretical insights and practical strategies for advancing AI to a stage where it can genuinely innovate, contributing meaningfully to science, technology, and the arts.
Digitalization is a crucial characteristic of the current era, and green innovation has become one of the necessary pathways for enterprises to achieve sustainable development. Based on financial and annual report data of Chinese A-share listed companies from 2010 to 2019, this paper constructs indicators of corporate digital transformation and examines the impact of corporate digital transformation on green innovation and its underlying mechanisms. The results show that corporate digital transformation can promote corporate green innovation output, with its sustained future impact exhibiting a marginally decreasing trend. In terms of the impact mechanism, digital transformation can enhance corporate green innovation output by increasing corporate R&D investment and strengthening environmental management. Heterogeneity analysis reveals that digital transformation has a more pronounced promoting effect on green innovation output for small and medium-sized enterprises and those in technology-intensive industries. To improve the green innovation incentive effect of digital transformation, enterprises should formulate long-term strategies and continuously strengthen policy regulation and incentives.
Vinícius de Santi Phelippe Nunes, Aparecido Vicente da Cruz Junior
A busca por métodos construtivos mais eficientes, sustentáveis e industrializados tem impulsionado a adoção de alternativas ao modelo tradicional de concreto armado. Entre essas soluções, destaca-se o sistema Steel Frame, amplamente utilizado em países desenvolvidos, mas ainda pouco explorado no Brasil. Caracterizado pela leveza dos componentes, rapidez na execução e menor geração de resíduos, o sistema representa uma resposta promissora aos desafios da construção civil contemporânea. Neste contexto, este artigo analisa tecnicamente a viabilidade do Steel Frame no cenário brasileiro por meio de uma abordagem comparativa com o sistema convencional, através do estudo de caso em uma residência unifamiliar, com análise estrutural, levantamento de cargas atuantes, orçamentos e cronogramas de execução. Os resultados revelam que, apesar do custo inicial mais elevado, o Steel Frame apresenta vantagens em termos de tempo de obra, desempenho técnico e sustentabilidade, consolidando-se como alternativa viável para edificações de pequeno porte.
The Indian biscuit industry, one of the fastest-growing segments in the processed food sector, has undergone significant transformation in recent decades. From traditional glucose biscuits to premium and functional baked goods, this evolution has been fueled by technological advancements, shifting consumer preferences, and innovations in process optimization. This article provides a historical overview of biscuit manufacturing in India, analyses current market trends, and explores the factors contributing to the industry's growth, including new product development and diversification into health-oriented and indulgent segments. Special attention is given to industrial observations made during hands-on training, with a focus on the analysis of process parameters such as dough consistency, baking conditions, and filling control in high-efficiency production lines. The integration of automation, quality assurance systems, and ingredient innovation is also discussed as key drivers of modern biscuit production. This study concludes that aligning consumer demands with technological upgrades is essential for sustained competitiveness and future innovation in India’s biscuit sector.
Yurii Dreval, Svitlana Zaika, Rastislav Tešovič
et al.
This study examines the pivotal role of the International Labour Organization (ILO) in advancing the United Nations Sustainable Development Goals (SDGs), with a particular focus on promoting decent work and ensuring safe and secure working conditions. The article explores the mechanisms of international regulation of social and labour relations, highlighting both the historical and contemporary contributions of the ILO to global labour standards. It evaluates the transformative impact of the COVID-19 pandemic on progress toward achieving these goals, emphasising the emergent challenges and the need for innovative measures to enhance occupational safety and health. The study also analyses the ILO's historical legacy, which has shaped the international labour landscape over the past century, and discusses its current initiatives to address modern labour challenges. Special attention is given to the interconnection between economic growth, financial sustainability, and the implementation of international labour standards. By exploring specific strategies for improving workplace conditions and fostering financial resilience, the research presents actionable recommendations for global policymakers, organisations, and stakeholders. The findings underscore the importance of adopting a multidimensional and cooperative approach to resolving global social and labour issues. The study concludes that achieving the SDGs - particularly Goal 8 (Decent Work and Economic Growth) - requires strengthened international collaboration and a shared commitment to aligning labour policies with sustainable development objectives. This comprehensive examination of the ILO's role provides valuable insights into its strategic importance in ensuring equitable and sustainable progress in labour relations worldwide.
This article explores the historical evolution, current transformation, and strategic importance of the school subject "Technologies" in Ukraine, particularly under the conditions of full-scale war. The study traces the transition from the Soviet-era utilitarian "labour training" course to a multidimensional, interdisciplinary field integrated within the New Ukrainian School (NUS) reform. The shift in educational paradigm is examined — from manual tasks to a focus on critical thinking, creativity, digital literacy, and entrepreneurial competencies. A special emphasis is placed on project-based learning, the application of digital tools (3D modeling, Arduino, Canva, Scratch, Google Workspace), and the implementation of STEAM approaches. The article highlights how technology lessons facilitate interdisciplinary collaboration with computer science, mathematics, biology, and arts education. The impact of the Russian invasion is analyzed in detail, illustrating how the subject has become not only an educational tool but also a psychosocial support mechanism, offering students a sense of stability, agency, and purpose through practical activities like candle making, sewing kits, and digital presentations. Key systemic challenges are discussed: unequal access to modern equipment, insufficient teacher training in digital and project-oriented methodologies, the lack of clear models for interdisciplinary integration, and outdated teacher certification procedures that fail to reflect 21st-century educational realities. Despite the difficulties, the Ukrainian education system continues to adapt and evolve. The article argues that the "Technologies" subject serves as a platform for resilience, innovation, and hands-on learning – forming an essential component of a future-oriented, inclusive, and socially relevant education system in times of both crisis and recovery.
The recent success of Large Language Models (LLMs) signifies an impressive stride towards artificial general intelligence. They have shown a promising prospect in automatically completing tasks upon user instructions, functioning as brain-like coordinators. The associated risks will be revealed as we delegate an increasing number of tasks to machines for automated completion. A big question emerges: how can we make machines behave responsibly when helping humans automate tasks as personal copilots? In this paper, we explore this question in depth from the perspectives of feasibility, completeness and security. In specific, we present Responsible Task Automation (ResponsibleTA) as a fundamental framework to facilitate responsible collaboration between LLM-based coordinators and executors for task automation with three empowered capabilities: 1) predicting the feasibility of the commands for executors; 2) verifying the completeness of executors; 3) enhancing the security (e.g., the protection of users' privacy). We further propose and compare two paradigms for implementing the first two capabilities. One is to leverage the generic knowledge of LLMs themselves via prompt engineering while the other is to adopt domain-specific learnable models. Moreover, we introduce a local memory mechanism for achieving the third capability. We evaluate our proposed ResponsibleTA on UI task automation and hope it could bring more attentions to ensuring LLMs more responsible in diverse scenarios.
Large-scale software development has become a highly collaborative and geographically distributed endeavour, especially in open-source software development ecosystems and their associated developer communities. It has given rise to modern development processes (e.g., pull-based development) that involve a wide range of activities such as issue and bug handling, code reviewing, coding, testing, and deployment. These often very effort-intensive activities are supported by a wide variety of tools such as version control systems, bug and issue trackers, code reviewing systems, code quality analysis tools, test automation, dependency management, and vulnerability detection tools. To reduce the complexity of the collaborative development process, many of the repetitive human activities that are part of the development workflow are being automated by CI/CD tools that help to increase the productivity and quality of software projects. Social coding platforms aim to integrate all this tooling and workflow automation in a single encompassing environment. These social coding platforms gave rise to the emergence of development bots, facilitating the integration with external CI/CD tools and enabling the automation of many other development-related tasks. GitHub, the most popular social coding platform, has introduced GitHub Actions to automate workflows in its hosted software development repositories since November 2019. This chapter explores the ecosystems of development bots and GitHub Actions and their interconnection. It provides an extensive survey of the state-of-the-art in this domain, discusses the opportunities and threats that these ecosystems entail, and reports on the challenges and future perspectives for researchers as well as software practitioners.
Within recent times, cybercriminals have curated a variety of organised and resolute cyber attacks within a range of cyber systems, leading to consequential ramifications to private and governmental institutions. Current security-based automation and orchestrations focus on automating fixed purpose and hard-coded solutions, which are easily surpassed by modern-day cyber attacks. Research within Automated Cyber Defence will allow the development and enabling intelligence response by autonomously defending networked systems through sequential decision-making agents. This article comprehensively elaborates the developments within Automated Cyber Defence through a requirement analysis divided into two sub-areas, namely, automated defence and attack agents and Autonomous Cyber Operation (ACO) Gyms. The requirement analysis allows the comparison of automated agents and highlights the importance of ACO Gyms for their continual development. The requirement analysis is also used to critique ACO Gyms with an overall aim to develop them for deploying automated agents within real-world networked systems. Relevant future challenges were addressed from the overall analysis to accelerate development within the area of Automated Cyber Defence.
Jimmi Normann Kristiansen, Catarina Batista, Tuuli Maria Utriainen
This special issue presents 4 selected papers that have an emphasis on either the antecedents or provide concrete cases of impact innovation. Across the papers, the authors approach the topic of impact innovation from distinct angles, from measures of personal innovativeness to the power of physical teamwork, to the purpose of prototyping and entrepreneurial attitudes. This serves to demonstrate that innovation is not a linear process but rather a complex phenomenon that can be studied from a multitude of technical and social perspectives.
Abstract In the field of industrial design and manufacture, computer‐supported collaborative work (CSCW) systems have been widely deployed for better teamwork. However, the traditional CSCW systems have a main drawback in effectively processing and utilising knowledge across different industrial workflows. To bridge this gap, we propose a framework for collaboration between members across the manufacturing value chains to increase efficiency and reduce duplication in team cooperation. The framework contains three parts, namely workflow, knowledge mining, and services. Specifically, the workflow part provides a collaborative environment for multiple users. The knowledge mining part, as the core of the framework, extracts in‐context knowledge from workflows. The part of services can interact with users with different users in each workflow, including information recommendation they need in the future or information retrieval they want to know from other workflows. Furthermore, we develop a prototype system for supporting multiple value chains collaboration to verify the effectiveness and efficiency of the framework.
Industry 4.0, as the most disruptive industrial revolution, is reshaping the industries by coupling the cyber to the physical systems. The current digitalization has its root in digitization introduced by Industry 3.0, where it made a foundation for gradual industrial migration to Industry 4.0. Respectively, Shipping 3.0 introduced automation and computerized systems onboard ships and paved the way for entering Shipping 4.0. In turn, the ultimate goal of Shipping 4.0 is full autonomy through the implementation of autonomous shipping. The introduction of autonomous shipping not only modifies the maritime workplaces but also changes the jobs’ definitions and the role of seafarers as the human element in the system. However, the journey to Shipping 4.0 will take the shipping industry to different steps before ships become fully autonomous. International Maritime Organization (IMO) defined these steps in four degrees from traditional to smart shipping. This implies that, at the same time, seafarers’ role and the required skills and competencies to gradually evolve with the ship's transition to the next degree. The review of literature about Industry 4.0 shows that so far, the focus of researchers and the industry is mainly on the innovation in technology and its implementation on ships. However, the role of the human and the cognitive human factor in the process is yet to be investigated. This paper aims to explore the effects of adaptation of digitalization in the shipping industry with a focus on the human element and cognitive human factor. The paper illustrates how the innovation and technological development of Industry 4.0 is changing the shipping industry and evolving human operators’ roles, responsibilities, and training needs.
The contribution presents results of the research focused on the adaptation of small and medium-sized enterprises (SMEs) in the service sector to technological, economic, social and environmental conditions of Industry and Society 4.0. The main goals of the research were the analysis and evaluation of the current state, preparedness, motivation and needs of SMEs in the sector of knowledge-intensive services for the timely, purposeful and effective implementation of Industry 4.0 methods and tools in the South-West region of the Czech Republic. The methodological approach is based on a mixed research strategy. Qualitative and quantitative methods such as desk research, explanatory and interpretation methods, questionnaire survey, semi-structured interview and evaluation of data and information were used to achieve the research goals. The results document the high current and future need to increase the knowledge and innovation potential of SMEs, the need for changes in the organisation and content of work, the need for changes in the competencies of employees in the context of robotisation, automation and digitisation of business processes. The results demonstrate barriers to access to sufficient internal and external financial resources, as well as a strong interest of SMEs in cooperation with the academic sector and regional authorities in the development of methodological tools for adaptation to new societal conditions and in the interest of the sustainable existence of these enterprises. The direction of future research is oriented towards the creation of a methodology for the adaptation of SMEs to the conditions of Industry and Society 4.0.
Faith Ibukun Babalola, Eseoghene Kokogho, Princess Eloho Odio
et al.
Audit quality remains a critical determinant of financial market stability, influencing investor confidence, corporate governance, and regulatory compliance. However, defining and assessing audit effectiveness in modern financial markets presents ongoing challenges due to evolving regulatory landscapes, technological disruptions, and market dynamics. This paper develops a conceptual framework for evaluating audit quality by integrating theoretical perspectives, key determinants, and emerging technological advancements. It explores the historical evolution of audit quality, the influence of regulatory and professional standards, and the role of ethics, independence, and professional skepticism in shaping audit effectiveness. The study identifies key drivers of audit effectiveness, including auditor competence, firm structure, regulatory oversight, and market forces, while highlighting emerging risks such as financial innovation, fraud, and digital disruptions. A robust conceptual framework is proposed, incorporating qualitative and quantitative performance metrics, focusing on artificial intelligence, big data, and automation as transformative forces in audit practices. The findings underscore the need for a holistic approach to audit quality assessment that aligns with investor and regulatory expectations while addressing gaps in existing models. The paper concludes with implications for regulators, audit firms, and policymakers, emphasizing the importance of strengthening oversight mechanisms, enhancing auditor independence, leveraging technology, and promoting transparency in financial reporting. Recommendations for improving audit effectiveness include regulatory innovations, ethical governance, technological integration, and specialized auditor training to adapt to the complexities of evolving financial markets. By advancing a multidimensional framework, this research contributes to the ongoing discourse on audit quality, offering practical insights for enhancing the reliability and integrity of financial audits in a rapidly changing environment.
Michael Desmond, Evelyn Duesterwald, Vatche Isahagian
et al.
Most business process automation is still developed using traditional automation technologies such as workflow engines. These systems provide domain specific languages that require both business knowledge and programming skills to effectively use. As such, business users often lack adequate programming skills to fully leverage these code oriented environments. We propose a paradigm for the construction of business automations using natural language. The approach applies a large language model to translate business rules and automations described in natural language, into a domain specific language interpretable by a business rule engine. We compare the performance of various language model configurations, across various target domains, and explore the use of constrained decoding to ensure syntactically correct generation of output.
Objectives: An SLR is presented focusing on text mining based automation of SLR creation. The present review identifies the objectives of the automation studies and the aspects of those steps that were automated. In so doing, the various ML techniques used, challenges, limitations and scope of further research are explained. Methods: Accessible published literature studies that primarily focus on automation of study selection, study quality assessment, data extraction and data synthesis portions of SLR. Twenty-nine studies were analyzed. Results: This review identifies the objectives of the automation studies, steps within the study selection, study quality assessment, data extraction and data synthesis portions that were automated, the various ML techniques used, challenges, limitations and scope of further research. Discussion: We describe uses of NLP/TM techniques to support increased automation of systematic literature reviews. This area has attracted increase attention in the last decade due to significant gaps in the applicability of TM to automate steps in the SLR process. There are significant gaps in the application of TM and related automation techniques in the areas of data extraction, monitoring, quality assessment and data synthesis. There is thus a need for continued progress in this area, and this is expected to ultimately significantly facilitate the construction of systematic literature reviews.