As industrial manufacturing scales, automating fine-grained product image analysis has become critical for quality control. However, existing approaches are hindered by limited dataset coverage and poor model generalization across diverse and complex anomaly patterns. To address these challenges, we introduce MAU-Set, a comprehensive dataset for Multi-type industrial Anomaly Understanding. It spans multiple industrial domains and features a hierarchical task structure, ranging from binary classification to complex reasoning. Alongside this dataset, we establish a rigorous evaluation protocol to facilitate fair and comprehensive model assessment. Building upon this foundation, we further present MAU-GPT, a domain-adapted multimodal large model specifically designed for industrial anomaly understanding. It incorporates a novel AMoE-LoRA mechanism that unifies anomaly-aware and generalist experts adaptation, enhancing both understanding and reasoning across diverse defect classes. Extensive experiments show that MAU-GPT consistently outperforms prior state-of-the-art methods across all domains, demonstrating strong potential for scalable and automated industrial inspection.
Abstract Humanity faces an unprecedented challenge in the necessity to rapidly change behaviors across various life domains to address multiple environmental crises, such as climate change, pollution, and biodiversity loss. This includes the behavior of individuals at work and within organizations. Industrial and organizational (I-O) psychology is uniquely positioned to provide evidence-based recommendations for changing organizational decision-making and behavior toward greater environmental sustainability. Although a substantial body of research on this topic has emerged over the past decade, the discipline has yet to realize its full potential because the topic is currently not prioritized and the practical and societal impact of previous research is limited. This article aims to propel research on environmental sustainability at work forward. To do so, it (a) outlines the interconnections between organizations and environmental sustainability; (b) portrays previous research efforts on environmental sustainability at work, resulting in an integrative conceptual framework across micro, meso, macro, and magno levels; and (c) provides actionable recommendations for high-impact future I-O psychology research and practice related to environmental sustainability. Following an “impact-first” rationale, we identified 10 areas for future research across the four levels of the conceptual framework. For each area, we present relevant theoretical perspectives, methodological approaches, and connections to related disciplines. Finally, we provide suggestions for effective science–practice transfer. Overall, the article seeks to spark discussion on this crucial topic within the community and to inspire I-O psychology researchers and practitioners to contribute to environmental sustainability.
This study proposes a new model based on a classic S-curve that describes deployment and stabilization at maximum capacity. In addition, the model extends to the post-growth plateau, where technological capacity is renewed according to the distribution of equipment lifespans. We obtain two qualitatively different results. In the case of "fast" deployment, characterized by a short deployment time in relation to the average equipment lifetime, production is subject to significant oscillations. In the case of "slow" deployment, production increases monotonically until it reaches a renewal plateau. These results are counterintuitively validated by two case studies: nuclear power plants as a fast deployment and smartphones as a slow deployment. These results are important for long-term industrial planning, as they enable us to anticipate future business cycles. Our study demonstrates that business cycles can originate endogenously from industrial dynamics of installation and renewal, contrasting with traditional views attributing fluctuations to exogenous macroeconomic factors. These endogenous cycles interact with broader trends, potentially being modulated, amplified, or attenuated by macroeconomic conditions. This dynamic of deployment and renewal is relevant for long-life infrastructure technologies, such as those supporting the renewable energy sector and has major policy implications for industry players.
Rinieri Lorenzo, Gori Giacomo, Melis Andrea
et al.
The OPC UA protocol is becoming the de facto standard for Industry 4.0 machine-to-machine communication. It stands out as one of the few industrial protocols that provide robust security features designed to prevent attackers from manipulating and damaging critical infrastructures. However, prior works showed that significant challenges still exists to set up secure OPC UA deployments in practice, mainly caused by the complexity of certificate management in industrial scenarios and the inconsistent implementation of security features across industrial OPC UA devices. In this paper, we present Pk-IOTA, an automated solution designed to secure OPC UA communications by integrating programmable data plane switches for in-network certificate validation and leveraging the IOTA Tangle for decen- tralized certificate distribution. Our evaluation is performed on a physical testbed representing a real-world industrial scenario and shows that Pk-IOTA introduces a minimal overhead while providing a scalable and tamper-proof mechanism for OPC UA certificate management.
Stefano Scanzio, Paolo Campagnale, Pietro Chiavassa
et al.
QR codes are nowadays customarily used for embedding static data such as web hyperlinks or plain text. The sQRy technology (executable QR codes) permits to embed executable programs in QR codes, enabling people to interact with them even without an internet connection. In this work we present QRmap, a specific dialect that permits the inclusion of geographic maps in sQRy and supports interaction with the user to provide indications to reach the destination of interest. The QRmap technology facilitates navigation in large industrial plants where internet connectivity is absent, due to either environmental limitations or company policies. The proposed technology can have interesting applications in non-industrial contexts as well.
In this paper, a new model-free anomaly detection framework is proposed for time-series induced by industrial dynamical systems.The framework lies in the category of conventional approaches which enable appealing features such as a learning with reduced amount of training data, a high potential for explainability as well as a compatibility with incremental learning mechanism to incorporate operator feedback after an alarm is raised and analyzed. Although these are crucial features towards acceptance of data-driven solutions by industry, they are rarely considered in the comparisons that generally almost exclusively focus on performance metrics. Moreover, the features engineering step involved in the proposed framework is inspired by the time-series being implicitly governed by physical laws as it is generally the case in industrial time-series. Two examples are given to assess the efficiency of the proposed approach.
The increasing demand for highly customized products, as well as flexible production lines, can be seen as trigger for the "fourth industrial revolution", referred to as "Industrie 4.0". Current systems usually rely on wire-line technologies to connect sensors and actuators. To enable a higher flexibility such as moving robots or drones, these connections need to be replaced by wireless technologies in the future. Furthermore, this facilitates the renewal of brownfield deployments to address Industrie 4.0 requirements. This paper proposes representative use cases, which have been examined in the German Tactile Internet 4.0 (TACNET 4.0) research project. In order to analyze these use cases, this paper identifies the main challenges and requirements of communication networks in Industrie 4.0 and discusses the applicability of 5th generation wireless communication systems (5G).
The possibility of block chain innovation revolutionizing business operations and interpersonal interactions in Industry 4.0 is becoming more widely acknowledged. Industry 4.0 and the Industrial Internet of Things (IoT) are among the new application fields. As a result, the purpose of this article is to investigate the block chain applications that are already being used in IoT and Industry 4.0. In particular, it looks at current research trends in various IoT applications, addressing problems, concerns, and potential future uses of integrating block chain technology. This article also includes a thorough discussion of the key elements of block chain databases, including Merkle trees, transaction management, sharding, long-term memory, and short-term memory. In order to do this, more than 46 pertinent primary research that have been published in reputable journals have been chosen for additional examination. The workflow of a block chain network utilizing IoT is also demonstrated, demonstrating how IoT devices communicate with one another and how they contribute to the network's overall operation. The taxonomy diagram below serves to illustrate the contribution.
André Luiz Buarque Vieira e Silva, Francisco Simões, Danny Kowerko
et al.
Within (semi-)automated visual industrial inspection, learning-based approaches for assessing visual defects, including deep neural networks, enable the processing of otherwise small defect patterns in pixel size on high-resolution imagery. The emergence of these often rarely occurring defect patterns explains the general need for labeled data corpora. To alleviate this issue and advance the current state of the art in unsupervised visual inspection, this work proposes a DifferNet-based solution enhanced with attention modules: AttentDifferNet. It improves image-level detection and classification capabilities on three visual anomaly detection datasets for industrial inspection: InsPLAD-fault, MVTec AD, and Semiconductor Wafer. In comparison to the state of the art, AttentDifferNet achieves improved results, which are, in turn, highlighted throughout our quali-quantitative study. Our quantitative evaluation shows an average improvement - compared to DifferNet - of 1.77 +/- 0.25 percentage points in overall AUROC considering all three datasets, reaching SOTA results in InsPLAD-fault, an industrial inspection in-the-wild dataset. As our variants to AttentDifferNet show great prospects in the context of currently investigated approaches, a baseline is formulated, emphasizing the importance of attention for industrial anomaly detection both in the wild and in controlled environments.
Industrial wireless sensor networks enable real-time data collection, analysis, and control by interconnecting diverse industrial devices. In these industrial settings, power outlets are not always available, and reliance on battery power can be impractical due to the need for frequent battery replacement or stringent safety regulations. Battery-less energy harvesters present a suitable alternative for powering these devices. However, these energy harvesters, equipped with supercapacitors instead of batteries, suffer from intermittent on-off behavior due to their limited energy storage capacity. As a result, they struggle with extended or frequent energy-consuming phases of multi-hop network formation, such as network joining and synchronization. To address these challenges, our work proposes three strategies for integrating battery-less energy harvesting devices into industrial multi-hop wireless sensor networks. In contrast to other works, our work prioritizes the mitigation of intermittency-related issues, rather than focusing solely on average energy consumption, as is typically the case with battery-powered devices. For each of the proposed strategies, we provide an in-depth discussion of their suitability based on several critical factors, including the type of energy source, storage capacity, device mobility, latency, and reliability.
Yuri Chervonyi, Praneet Dutta, Piotr Trochim
et al.
We present a hybrid industrial cooling system model that embeds analytical solutions within a multi-physics simulation. This model is designed for reinforcement learning (RL) applications and balances simplicity with simulation fidelity and interpretability. The model's fidelity is evaluated against real world data from a large scale cooling system. This is followed by a case study illustrating how the model can be used for RL research. For this, we develop an industrial task suite that allows specifying different problem settings and levels of complexity, and use it to evaluate the performance of different RL algorithms.
When changes are performed on an automated production system (aPS), new faults can be accidentally introduced in the system, which are called regressions. A common method for finding these faults is regression testing. In most cases, this regression testing process is performed under high time pressure and on-site in a very uncomfortable environment. Until now, there is no automated support for finding and prioritizing system test cases regarding the fully integrated aPS that are suitable for finding regressions. Thus, the testing technician has to rely on personal intuition and experience, possibly choosing an inappropriate order of test cases, finding regressions at a very late stage of the test run. Using a suitable prioritization, this iterative process of finding and fixing regressions can be streamlined and a lot of time can be saved by executing test cases likely to identify new regressions earlier. Thus, an approach is presented in this paper that uses previously acquired runtime data from past test executions and performs a change identification and impact analysis to prioritize test cases that have a high probability to unveil regressions caused by side effects of a system change. The approach was developed in cooperation with reputable industrial partners active in the field of aPS engineering, ensuring a development in line with industrial requirements. An industrial case study and an expert evaluation were performed, showing promising results.
Industrial Internet of Things (IIoT) revolutionizes the future manufacturing facilities by integrating the Internet of Things technologies into industrial settings. With the deployment of massive IIoT devices, it is difficult for the wireless network to support the ubiquitous connections with diverse quality-of-service (QoS) requirements. Although machine learning is regarded as a powerful data-driven tool to optimize wireless network, how to apply machine learning to deal with the massive IIoT problems with unique characteristics remains unsolved. In this paper, we first summarize the QoS requirements of the typical massive non-critical and critical IIoT use cases. We then identify unique characteristics in the massive IIoT scenario, and the corresponding machine learning solutions with its limitations and potential research directions. We further present the existing machine learning solutions for individual layer and cross-layer problems in massive IIoT. Last but not the least, we present a case study of massive access problem based on deep neural network and deep reinforcement learning techniques, respectively, to validate the effectiveness of machine learning in massive IIoT scenario.
Industry 4.0 describes an adaptive and changeable production, where its factory cells have to be reconfigured at very short intervals, e.g. after each workpiece. Furthermore, this scenario cannot be realized with traditional devices, such as programmable logic controllers. Here the use of well-proven technologies of the information technology are conquering the production hall (IT-OT convergence). Therefore, both virtualization and novel communication technologies are being introduced in the field of industrial automation. In addition, these technologies are seen as the key for facilitating various emerging use cases. However, it is not yet clear whether each of the dedicated hardware and software components, which have been developed for specific control tasks and have performed well over decades, can be upgraded without major adjustments. In this paper, we examine the opportunities and challenges of hardware and operating system-level virtualization based on the stringent requirements imposed by industrial applications. For that purpose, benchmarks for different virtualization technologies are set by determining their computational and networking overhead, configuration effort, accessibility, scalability, and security.
Florian Fischer, Matthias Niedermaier, Thomas Hanka
et al.
More and more industrial devices are connected to IP-based networks, as this is essential for the success of Industry 4.0. However, this interconnection also results in an increased attack surface for various network-based attacks. One of the easiest attacks to carry out are DoS attacks, in which the attacked target is overloaded due to high network traffic and corresponding CPU load. Therefore, the attacked device can no longer provide its regular services. This is especially critical for devices, which perform real-time operations in industrial processes. To protect against DoS attacks, there is the possibility of throttling network traffic at the perimeter, e.g. by a firewall, to develop robust device architectures. In this paper, we analyze various concepts for secure device architectures and compare them with regard to their robustness against DoS attacks. Here, special attention is paid to how the control process of an industrial controller behaves during the attack. For this purpose, we compare different schedulers on single-core and dual-core Linux-based systems, as well as a heterogeneous multi-core architecture under various network loads and additional system stress.
Alfonso Di Pace, Giuseppe Fenza, Mariacristina Gallo
et al.
In the current industrial practices, the exponential growth in terms of availability and affordability of sensors, data acquisition systems, and computer networks forces factories to move toward implementing high integrating Cyber-Physical Systems (CPS) with production, logistics, and services. This transforms today's factories into Industry 4.0 factories with significant economic potential. Industry 4.0, also known as the fourth Industrial Revolution, levers on the integration of cyber technologies, the Internet of Things, and Services. This paper proposes an Augmented Reality (AR)-based system that creates a Cognition Level that integrates existent Manufacturing Execution Systems (MES) to CPS. The idea is to highlight the opportunities offered by AR technologies to CPS by describing an application scenario. The system, analyzed in a real factory, shows its capacity to integrate physical and digital worlds strongly. Furthermore, the conducted survey (based on the Situation Awareness Global Assessment Technique method) reveals significant advantages in terms of production monitoring, progress, and workers' Situation Awareness in general.
Gianmarco Altoè, Giulia Bertoldo, Claudio Zandonella Callegher
et al.
In the past two decades, psychological science has experienced an unprecedented replicability crisis which uncovered several issues. Among others, statistical inference is too often viewed as an isolated procedure limited to the analysis of data that have already been collected. We build on and further develop an idea proposed by Gelman and Carlin (2014) termed "prospective and retrospective design analysis". Rather than focusing only on the statistical significance of a result and on the classical control of type I and type II errors, a comprehensive design analysis involves reasoning about what can be considered a plausible effect size. Furthermore, it introduces two relevant inferential risks: the exaggeration ratio or Type M error (i.e., the predictable average overestimation of an effect that emerges as statistically significant), and the sign error or Type S error (i.e., the risk that a statistically significant effect is estimated in the wrong direction). Another important aspect of design analysis is that it can be usefully carried out both in the planning phase of a study and for the evaluation of studies that have already been conducted, thus increasing researchers' awareness during all phases of a research project. We use a familiar example in psychology where the researcher is interested in analyzing the differences between two independent groups. We examine the case in which the plausible effect size is formalized as a single value, and propose a method in which uncertainty concerning the magnitude of the effect is formalized via probability distributions. Through several examples, we show that even though a design analysis requires big effort, it has the potential to contribute to planning more robust and replicable studies. Finally, future developments in the Bayesian framework are discussed.