Hasil untuk "Industrial safety. Industrial accident prevention"

Dalibor Sain, Thomas Rosenstatter, Olaf Saßnick et al.

The increased connectivity of industrial networks has led to a surge in cyberattacks, emphasizing the need for cybersecurity measures tailored to the specific requirements of industrial systems. Modern Industry 4.0 technologies, such as OPC UA, offer enhanced resilience against these threats. However, widespread adoption remains limited due to long installation times, proprietary technology, restricted flexibility, and formal process requirements (e.g. safety certifications). Consequently, many systems do not yet implement these technologies, or only partially. This leads to the challenge of dealing with so-called brownfield systems, which are often placed in isolated security zones to mitigate risks. However, the need for data exchange between secure and insecure zones persists. This paper reviews existing solutions to address this challenge by analysing their approaches, advantages, and limitations. Building on these insights, we identify three key concepts, evaluate their suitability and compatibility, and ultimately introduce the SigmaServer, a novel TCP-level aggregation method. The developed proof-of-principle implementation is evaluated in an operational technology (OT) testbed, demonstrating its applicability and effectiveness in bridging secure and insecure zones.

Xu Huang, Hao Zhang, Zhifang Fan et al.

As industrial recommender systems enter a scaling-driven regime, Transformer architectures have become increasingly attractive for scaling models towards larger capacity and longer sequence. However, existing Transformer-based recommendation models remain structurally fragmented, where sequence modeling and feature interaction are implemented as separate modules with independent parameterization. Such designs introduce a fundamental co-scaling challenge, as model capacity must be suboptimally allocated between dense feature interaction and sequence modeling under a limited computational budget. In this work, we propose MixFormer, a unified Transformer-style architecture tailored for recommender systems, which jointly models sequential behaviors and feature interactions within a single backbone. Through a unified parameterization, MixFormer enables effective co-scaling across both dense capacity and sequence length, mitigating the trade-off observed in decoupled designs. Moreover, the integrated architecture facilitates deep interaction between sequential and non-sequential representations, allowing high-order feature semantics to directly inform sequence aggregation and enhancing overall expressiveness. To ensure industrial practicality, we further introduce a user-item decoupling strategy for efficiency optimizations that significantly reduce redundant computation and inference latency. Extensive experiments on large-scale industrial datasets demonstrate that MixFormer consistently exhibits superior accuracy and efficiency. Furthermore, large-scale online A/B tests on two production recommender systems, Douyin and Douyin Lite, show consistent improvements in user engagement metrics, including active days and in-app usage duration.

Tarek Abu Selo, Zahid Hussain, Qinaat Hussain et al.

Simulation sickness (SS) remains a challenge in virtual reality (VR) applications, especially in pedestrian safety research. This study investigates SS symptoms in VR environments, focusing on gender differences, exposure time, and user perceptions. A total of 145 participants were exposed to two VR pedestrian scenarios: a crosswalk and a sidewalk. The Simulator Sickness Questionnaire (SSQ) was used to assess symptoms of nausea, oculomotor disturbance, and disorientation. Results showed that female participants reported significantly higher SS symptoms than males, with the sidewalk scenario inducing greater overall SS. Additionally, perceived realism in the VR environment was associated with reduced symptoms, while perceived disengagement led to increased discomfort. These findings highlight the importance of user perceptions in mitigating SS and suggest that VR scenarios should be designed with attention to gender differences and environmental realism to improve user experience and safety.

Xiaoran Yang, Yuyang Du, Kexin Chen et al.

As Industry 4.0 progresses, flexible manufacturing has become a cornerstone of modern industrial systems, with equipment automation playing a pivotal role. However, existing control software for industrial equipment, typically reliant on graphical user interfaces (GUIs) that require human interactions such as mouse clicks or screen touches, poses significant barriers to the adoption of code-based equipment automation. Recently, Large Language Model-based General Computer Control (LLM-GCC) has emerged as a promising approach to automate GUI-based operations. However, industrial settings pose unique challenges, including visually diverse, domain-specific interfaces and mission-critical tasks demanding high precision. This paper introduces IndusGCC, the first dataset and benchmark tailored to LLM-GCC in industrial environments, encompassing 448 real-world tasks across seven domains, from robotic arm control to production line configuration. IndusGCC features multimodal human interaction data with the equipment software, providing robust supervision for GUI-level code generation. Additionally, we propose a novel evaluation framework with functional and structural metrics to assess LLM-generated control scripts. Experimental results on mainstream LLMs demonstrate both the potential of LLM-GCC and the challenges it faces, establishing a strong foundation for future research toward fully automated factories. Our data and code are publicly available at: \href{https://github.com/Golden-Arc/IndustrialLLM}{https://github.com/Golden-Arc/IndustrialLLM.

Yuxi Liu, Yunfeng Ma, Yi Tang et al.

Industrial surface defect detection (SDD) is critical for ensuring product quality and manufacturing reliability. Due to the diverse shapes and sizes of surface defects, SDD faces two main challenges: intraclass difference and interclass similarity. Existing methods primarily utilize manually designed models, which require extensive trial and error and often struggle to address both challenges effectively. To overcome this, we propose AutoNAD, an automated neural architecture design framework for SDD that jointly searches over convolutions, transformers, and multi-layer perceptrons. This hybrid design enables the model to capture both fine-grained local variations and long-range semantic context, addressing the two key challenges while reducing the cost of manual network design. To support efficient training of such a diverse search space, AutoNAD introduces a cross weight sharing strategy, which accelerates supernet convergence and improves subnet performance. Additionally, a searchable multi-level feature aggregation module (MFAM) is integrated to enhance multi-scale feature learning. Beyond detection accuracy, runtime efficiency is essential for industrial deployment. To this end, AutoNAD incorporates a latency-aware prior to guide the selection of efficient architectures. The effectiveness of AutoNAD is validated on three industrial defect datasets and further applied within a defect imaging and detection platform. Code is available at https://github.com/Yuxi104/AutoNAD.

Tho Bella Dinh-Zarr, Ray Shuey, Lori Mooren

Amity Latham, Zoran Najdovski, Rebecca Bartel et al.

This project combined systems engineers, farm safety researchers, work health and safety inspectorate and policymakers with the aim of designing a way in which to reduce fatal farm injury caused by run-overs and roll-overs by tractors and side-by-side vehicles. The team made comparisons between farm machinery and powered mobile plant that is used in the industrial manufacturing, warehousing and logistics, mining, and construction sectors. Current and emerging safety technologies and engineering solutions were collated. Safety standards, legislated engineering controls, retrofit designs, and known ways in which farmers’ workaround safety features were considered. These elements were used as criteria to propose a way to resolve which safety technologies or engineering controls should be recommended for aftermarket retrofitting or incorporated at the original equipment manufacturer design stage. The concept of measuring safety effectiveness to prevent fatal farm injury emerged. This developed into a score sheet and a corresponding matrix to highlight engineering strength and industry acceptance. The project resulted in the conceptual design of the agricultural safety effectiveness score (ASES). The next phase is a multi-stakeholder validation process and a protocol for the scoring system. It requires a hypothesis to test the theory that when safety technologies and engineering solutions are mature in other industries or if they are associated with agricultural productivity gains, their adoption into the agricultural sector is more likely, which in turn will reduce the incidence of tractor and side-by-side run-overs and roll-overs on farms.

Volodymyr Tymchuk

Вирішення сукупності проблем щодо забезпечення надійного та безперервного функціонування й ефективного управління складними організаційними структурами, що володіють повнотою незалежності, потребує нових концептуальних і методологічних рішень. Особливо значущим це постає в секторі безпеки та оборони України в умовах здійснення Українською державою відсічі збройної агресії російської федерації, оскільки, апріорі, цей сектор передбачає і вимагає залучення усіх можливих ресурсів і можливостей, у тому числі й передових, без достатньої апробації та досвіду застосування, для своєчасного й ефективного виконання місій і завдань. У світі, теоретичною основою для взаємодії різних систем і організацій, незалежно від рівня, є науковий і методологічний апарат такої галузі як теорія систем із систем (system-of-systems (SoS)). Їх реалізації мають практичне застосування і, на сьогодні, та або інша система із систем призначена або для найвищого органу управління держави, або для міжнаціональної коаліційної діяльності, або для організації з питань безпеки, оборони, бойового застосування на всіх рівнях військового управління. Загалом, у цій особливій і складній структурі, якою є система із систем, або, з іншого погляду, системи у системі, нові можливості виникають завдяки взаємодії окремих незалежних повнофункціональних систем, якими можуть керувати різні команди операторів. В таких реаліях кожна система із систем буде унікальною, що визначається її особливою архітектурою, яка сприйнятлива до змін і еволюцій. Методологічні основи теорії систем із систем є маловідомими серед вітчизняних науковців. Водночас практика національного спротиву проявляє конкретні сфери, у яких їх застосування є доречним, потрібним, своєчасним і ефективним. Тож, метою статті є розроблення інформаційно-аналітичних основ мілітарних систем із систем, їх класифікації на основі емерджетних, еволюційних й інших властивостей із онтологією узагальнених архітектур для впровадження у сектор безпеки і оборони та розвитку його спроможностей. Під час написання статті застосовано пошуковий метод, метод відбору, упорядкування та аналізу джерел, а також метод онтологічного інжинірингу щодо розробки онтології для архітектури конкретного різновиду системи із систем. У результаті, в статті наведено архітектури, складові, призначення та ключові особливості мілітарних систем із систем Сполучених Штатів Америки, Китайської Народної Республіки і коаліції держав НАТО. Було введено критерій класифікації – відношення системи із систем до її еволюції, що дозволило здійснити розподіл на п’ять класів систем із систем: спроєктовані, перетворені, скомбіновані, зеволюційовані, пристосовані. Кожен клас описано своєю онтологією та проілюстровано прикладом. Розроблена класифікація дасть змогу порівняно легко відносити структурні взаємодії між незалежними системами та організаціями до одного з класів систем із систем, що в подальшому сприятиме застосуванню методології цієї теорії. Також у статті запропоновано низку термінів, що дасть змогу надалі обґрунтовано, з урахуванням регламентованих підходів до стандартизації, розробити необхідну термінологію для нової галузі, виокремленої із загальної теорії систем. Елементом наукової новизни є наведення складних систем в архітектурному рішенні та онтологічному апараті теорії систем із систем. Теоретичну значущість статті формують терміни і поняття, що доповнюють термінологічну область сфери системної інженерії у створенні передусім систем управління (військами, процесами, організаціями тощо). Практична значущість досліджень зводиться до, по-перше, опанування понятійного апарату держав-партнерів України в царині складних систем управління, в царині C4ISR, тим самим надаючи додаткове поле для взаємодії у питаннях безпеки та оборони, зокрема, щодо захисту України, і, по-друге, до можливості практичної реалізації конкретних систем із систем і систем у системі, цілеорієнтованих під певну військову місію або задачу чи діяльність.

Rakesh Gangadharaiah, Johnell O. Brooks, Lauren Mims et al.

This exploratory study investigated whether using the Pedals Emergency Stop© interactive driving simulator exercise improved the understanding and performance of emergency braking among novice teen drivers. Seventy-one high school driver education students (aged 15–19) participated. All of the teens completed the Pedals Emergency Stop© interactive exercise driving simulator task and then an on-road ABS exercise in a driver’s education vehicle; there was no control group. Students’ ability to complete the simulator-based emergency braking task increased from an initial passing rate of only 18.3% to a maximum of 81.7% by the end of the simulation exercise. A positive trend was observed over successive simulator trials, with the linear effect explaining 51.1% of the variance in emergency stopping “pass” rates using the simulator task. In addition, participants who passed more trials during the Pedals Emergency Stop© simulator exercise were 12.3% more likely to fully activate the ABS during the on-road emergency stop activity using the driver’s education vehicle. Post-study surveys revealed that 95% of the participants improved their understanding of ABS as a result of the simulation-based training, and 98% felt there was a positive impact from the driving simulation exercise on their real-world emergency braking capabilities. Participants highly endorsed the Pedals Emergency Stop© exercise for ABS education and refresher training, with a rating of 4.7 out of 5. This study emphasizes the potential benefits of incorporating simulator-based exercises into driver education and training, with the long-term goal of promoting safe driving behaviors and outcomes.

Sebastián Rojas-Innocenti, Enrique Baeyens, Alejandro Martín-Crespo et al.

The increasing integration of renewable energy sources into power systems is intensifying the demand for greater flexibility among industrial electricity consumers. However, operational constraints, production requirements, and market dynamics pose significant challenges to achieving optimal flexibility. This paper presents an enhanced mixed integer linear programming (MILP) model that directly optimizes electricity consumption flexibility in manufacturing plants. Unlike previous approaches, the proposed model determines optimal transactions with both day-ahead and intraday continuous electricity markets, while ensuring production continuity and adhering to plant-specific operational constraints. The methodology is validated through annual simulations of two real world industrial configurations, cement manufacturing and steel production, using 2023 market data. Comparative results highlight that the steel plant achieved average electricity cost savings through flexibility of 0.41 euro/MWh, whereas the cement plant achieved 0.24 euro/MWh, reflecting differences in storage capacities, production rates, and operational flexibility. A comprehensive sensitivity analysis further identifies key parameters affecting flexibility potential, such as the production to demand ratio, storage capacity, and minimum operation periods. The findings offer valuable insights for industrial operators aiming to reduce energy costs, enhance operational flexibility, and support the decarbonization of electricity systems.

Atif Hussain, Rana Rizwan

This paper argues for the strategic treatment of artificial intelligence as a key industry within broader industrial policy framework of Pakistan, underscoring the importance of aligning it with national goals such as economic resilience and preservation of autonomy. The paper starts with defining industrial policy as a set of targeted government interventions to shape specific sectors for strategic outcomes and argues for its application to AI in Pakistan due to its huge potential, the risks of unregulated adoption, and prevailing market inefficiencies. The paper conceptualizes AI as a layered ecosystem, comprising foundational infrastructure, core computing, development platforms, and service and product layers, supported by education, government policy, and research and development. The analysis highlights that AI sector of Pakistan is predominantly service oriented, with limited product innovation and dependence on foreign technologies, posing risks to economic independence, national security, and employment. To address these challenges, the paper recommends educational reforms, support for local AI product development, initiatives for indigenous cloud and hardware capabilities, and public-private collaborations on foundational models. Additionally, it advocates for public procurement policies and infrastructure incentives to foster local solutions and reduce reliance on foreign providers. This strategy aims to position Pakistan as a competitive, autonomous player in the global AI ecosystem.

Haili Sun, Yan Huang, Lansheng Han et al.

Multivariate Time Series (MTS) anomaly detection focuses on pinpointing samples that diverge from standard operational patterns, which is crucial for ensuring the safety and security of industrial applications. The primary challenge in this domain is to develop representations capable of discerning anomalies effectively. The prevalent methods for anomaly detection in the literature are predominantly reconstruction-based and predictive in nature. However, they typically concentrate on a single-dimensional instance level, thereby not fully harnessing the complex associations inherent in industrial MTS. To address this issue, we propose a novel self-supervised hierarchical contrastive consistency learning method for detecting anomalies in MTS, named HCL-MTSAD. It innovatively leverages data consistency at multiple levels inherent in industrial MTS, systematically capturing consistent associations across four latent levels-measurement, sample, channel, and process. By developing a multi-layer contrastive loss, HCL-MTSAD can extensively mine data consistency and spatio-temporal association, resulting in more informative representations. Subsequently, an anomaly discrimination module, grounded in self-supervised hierarchical contrastive learning, is designed to detect timestamp-level anomalies by calculating multi-scale data consistency. Extensive experiments conducted on six diverse MTS datasets retrieved from real cyber-physical systems and server machines, in comparison with 20 baselines, indicate that HCL-MTSAD's anomaly detection capability outperforms the state-of-the-art benchmark models by an average of 1.8\% in terms of F1 score.

Simone Tonini, Andrea Vandin, Francesca Chiaromonte et al.

We present a novel, simple and widely applicable semi-supervised procedure for anomaly detection in industrial and IoT environments, SAnD (Simple Anomaly Detection). SAnD comprises 5 steps, each leveraging well-known statistical tools, namely; smoothing filters, variance inflation factors, the Mahalanobis distance, threshold selection algorithms and feature importance techniques. To our knowledge, SAnD is the first procedure that integrates these tools to identify anomalies and help decipher their putative causes. We show how each step contributes to tackling technical challenges that practitioners face when detecting anomalies in industrial contexts, where signals can be highly multicollinear, have unknown distributions, and intertwine short-lived noise with the long(er)-lived actual anomalies. The development of SAnD was motivated by a concrete case study from our industrial partner, which we use here to show its effectiveness. We also evaluate the performance of SAnD by comparing it with a selection of semi-supervised methods on public datasets from the literature on anomaly detection. We conclude that SAnD is effective, broadly applicable, and outperforms existing approaches in both anomaly detection and runtime.

Yibing Wu, Shuguang Zhang, Xuan Zhang et al.

Xuesong Wang, Caiyang Ye, Mohammed Quddus et al.

Narelle Haworth, Amy Schramm, Divera Twisk

Mykhailo Rakushev, Yuriі Kravchenko , Roman Pantiushenko

У статті проведено аналіз супроводження розробником інформаційно-телекомунікаційної системи «Термінал» для виконання завдань збору, обробки, обміну та відображення геопросторової інформації видового спостереження. Розглянуто діяльність вітчизняної компанії «Товариство з обмеженою відповідальністю “УкрСпецСистемс”» щодо вдосконалення комп’ютерної програми «Термінал» з початку широкомасштабної збройної агресії російської федерації проти України – з лютого 2022 року до травня 2023 року. Проаналізовані основні напрями з удосконалення програмного забезпечення, а саме: режими роботи, використання електронних карт і відображення обстановки, політика безпеки та адміністрування, обробка матеріалів зйомки з безпілотних літальних апаратів, передавання й обробка відео, обмін геопросторовими даними з іншими системами військового призначення, ліцензування та політика оновлення програмного забезпечення. Наведено узагальнені часові показники виходу нових версій комп’ютерної програми «Термінал» та інтенсивності внесення змін до зазначеного програмного забезпечення. За результатами проведеного аналізу обґрунтовано висновок щодо суттєвого нарощування спроможностей інформаційно-телекомунікаційної системи «Термінал» для виконання завдань збору, обробки, обміну та відображення інформації видового спостереження, а саме: добутої безпілотними літальними апаратами, трансльованої стаціонарними камерами спостереження та отримуваної від космічних апаратів видового спостереження. Підтвердженням зазначеного є суттєве розширення мережі інформаційно-телекомунікаційної системи «Термінал» протягом 16 місяців широкомасштабної збройної агресії російської федерації проти України.

Leonie Boskeljon-Horst, Robert J. De Boer, Vincent Steinmetz et al.

Although rules support people while executing tasks, they are not the same as work-as-done. It can be impossible to follow the rules and finish the job at the same time. In this study, the objective is to better understand the stakes and interests that lie behind retaining gaps between work-as-prescribed and work-as-done, mapping the benefits and fears of noncompliance. The study was conducted along the vertical hierarchy of an operational flight squadron of the Royal Netherlands Air Force. We applied a qualitative survey research methodology using semi-structured interviews, complemented by an investigation of relevant documents. We found a public and political commitment to compliance made by the Dutch Department of Defence, which reinforces a cycle of issuing promises followed by pressure to keep the promise. This contradicts the found need for adaptation and freedom to use expertise. The official safety narrative seems to convey a hidden message—bad things happen to bad people, reminiscent of a bogeyman. One opportunity to resolve the situation is a doctrine change, changing prescriptive rules to guidelines.

Adi Saptari, Poh Kiat Ng, Michelle Junardi et al.

In manufacturing companies, manual material handling (MMH) involves lifting, pushing, pulling, carrying, moving, and lowering objects, which can lead to musculoskeletal disorders (MSDs) among workers, resulting in high labor costs due to excessive overtime incurred for manual product preparation. The aim of this study was to show how ergonomic measures were used to reduce the risk of MSDs and to reduce operating costs in the warehouse department of an oil and gas service company. A preliminary study using the Nordic Body Map survey showed that the workers experienced pain in various parts of the body, indicating the presence of MSDs. The researchers then used methods such as the Rapid Upper Limb Assessment (RULA), Rapid Entire Body Assessment (REBA), and National Institute for Occupational Safety and Health (NIOSH) assessments to verify whether the MMH activities had an acceptable level of risk. The results revealed that certain manual material handling (MMH) activities were assessed as low–very high risk, with RULA scores ranging from 3 to 7 and REBA scores ranging from 4 to 11. An immediate solution was to replace the manual process with a semi-automatic process using a vacuum lifter. A feasibility study was conducted using the net present value (NPV), internal rate of return (IRR), and payback period to justify the economic viability of the solution. The analysis indicated that implementing the vacuum lifter not only mitigated the risk of MSDs but also reduced the operating costs, demonstrating its viability and profitability. Overall, this study suggests that implementing a vacuum lifter as an assistive device in the warehouse would be a beneficial investment for both the workers and the company, improving both well-being and finances.

Steven Landgraf, Markus Hillemann, Moritz Aberle et al.

In many industrial processes, such as power generation, chemical production, and waste management, accurately monitoring industrial burner flame characteristics is crucial for safe and efficient operation. A key step involves separating the flames from the background through binary segmentation. Decades of machine vision research have produced a wide range of possible solutions, from traditional image processing to traditional machine learning and modern deep learning methods. In this work, we present a comparative study of multiple segmentation approaches, namely Global Thresholding, Region Growing, Support Vector Machines, Random Forest, Multilayer Perceptron, U-Net, and DeepLabV3+, that are evaluated on a public benchmark dataset of industrial burner flames. We provide helpful insights and guidance for researchers and practitioners aiming to select an appropriate approach for the binary segmentation of industrial burner flames and beyond. For the highest accuracy, deep learning is the leading approach, while for fast and simple solutions, traditional image processing techniques remain a viable option.