Disassembly plays a key role in the circular economy by enabling the recovery of valuable components from end-of-life products. However, the high level of manual labor involved, as well as the variability and uncertainty associated with end-of-life products, can render disassembly economically inefficient. To increase efficiency and resilience, disassembly systems must better respond to deviations observed during operations, such as damaged components.
Reactive disassembly planning can achieve this, but requires evaluating alternative disassembly plans based on their impact on overall system performance to select suitable plan adaptations in response to a given deviation. This evaluation should be data-driven and consider the current system state, making real-time simulation a suitable approach. Existing approaches either lack real-time data integration and near-real-time response capabilities or fail to provide the required scope and level of detail for modeling disassembly systems. To address this gap, this article introduces a real-time simulation framework for the operational evaluation of disassembly plans, comprising three modules for configuration, simulation, and evaluation. Given multiple alternative disassembly plans, the framework enables the automatic simulation of system behavior for each plan, based on the current state of the disassembly system. Based on the simulation data, alternatives are evaluated using a multi-criteria indicator-based approach that integrates activity-based costing. This enables the data-driven selection of the most suitable disassembly plan.
The framework is practically validated using an automotive case study in a learning factory, demonstrating its functionality and performance.

Medical device remanufacturing can significantly reduce the waste generation and greenhouse gas emissions caused by the healthcare sector. However, its implementation is hindered by regulatory uncertainty and fragmentation, as well as by the lack of concrete process-level requirements for remanufacturing. This paper presents a systematization of European regulations and international standards applicable to medical device remanufacturing. The systematization is based on literature and regulatory sources and comprises two parts. The first part covers general requirements related to the quality management system of a remanufacturer, including a risk management and post-market surveillance system, and technical documentation. The second part provides an in-depth overview of requirements and applicable harmonized standards for each operational step in the remanufacturing process specific to medical devices. The systematization offers guidance for the regulatory-compliant implementation of remanufacturing by providing a structured overview and supports the sector’s transition towards greater circularity.

The transition toward sustainable industrial systems requires robust methodologies capable of identifying inefficiencies embedded in regulated waste management practices. In pharmaceutical manufacturing, conservative classification rules adopted to ensure compliance with Good Manufacturing Practices (GMP) may lead to systematic over-classification of waste as hazardous. While precautionary in nature, this practice can generate unintended economic burdens and environmental impacts due to unnecessary high-intensity treatment routes such as hazardous incineration. This study proposes a repeatable and quantitative methodological framework to estimate hazardous waste misclassification rates and to evaluate their economic and environmental implications. The approach integrates: (i) process-level waste flow mapping; (ii) explicit separation between internal regulatory classification and effective contamination state; (iii) statistical sampling to estimate a misclassification rate with defined confidence intervals; and (iv) scenario-based comparison between current (AS-IS) and optimized (TO-BE) waste management routes. Economic impacts are calculated through differential treatment and logistics costs, while environmental effects are assessed using a simplified comparative life cycle approach. By integrating statistical auditing with sustainability assessment, the proposed method supports pharmaceutical manufacturers in reducing over-classification while maintaining regulatory compliance and improving overall environmental and economic performance.

Immersive videos are emerging as a powerful tool for transforming knowledge-intensive services into scalable, standardized, and internationally transferable formats. The collaborative project presented in this article investigates how realistic 180° video recordings can be used for training purposes to make learning processes—which have traditionally been tied to specific locations and trainers—more flexible and reproducible. The focus is particularly on the fields of occupational and cybersecurity, where the authentic depiction of critical situations can make a significant contribution to raising awareness, promoting safe behavior, and developing competencies.

The production of immersive videos creates a modular training program that can be delivered independently of physical presence, thereby enabling significant gains in productivity and efficiency. In addition, the format opens up new potential for internationalization: multilingual versions can be created without producing new content, as only audio and text components need to be localized. This reduces costs, facilitates the global distribution of training, and promotes the exportability of services.

The project results show that immersive video formats have great potential to standardize service processes, reduce variations in the quality of knowledge transfer, and expand the reach of training programs across national borders. The results thus contribute both to the further development of digital continuing education formats and to the strategic scaling of knowledge-intensive services in international markets.

Urban environments concentrate people, critical assets and interconnected infrastructures, creating conditions in which localised shocks can propagate rapidly across sectors and administrative boundaries. Social media and sensor-generated data streams provide near-real-time traces of such disruptions, yet extracting reliable signals and converting them into actionable urban-risk knowledge still presents considerable conceptual, technical and ethical challenges. This paper proposes a comprehensive methodology for mining event-related messages from multilingual social and open-source intelligence (OSINT) data streams, coupled with the construction of a context-sensitive urban-risk taxonomy. The approach combines top-down alignment with the United Nations Office for Disaster Risk Reduction (UNDRR) multi-hazard frameworks, including the Climate Resilience Classification Framework launched at COP28, with bottom-up, data-driven discovery from heterogeneous digital sources. We adopt a cyclical analytical framework articulated through six iterative phases (Observe, Orient, Curate, Model, Validate, Reflect) and apply it to two contrasting European case-study cities: Bratislava (Slovakia) and Odesa (Ukraine). The technical pipeline encompasses streaming ingestion, multilingual natural language processing, burst and topic-shift detection, BERT-based zero-shot classification, spatiotemporal clustering via DBSCAN, cross-source corroboration through graph matching, and human-in-the-loop validation with active learning. Hierarchical agglomerative clustering on event-embedding vectors extends the taxonomy to capture emergent risks absent from established catalogues, while a Git-backed JSON-LD versioning system ensures transparency and reproducibility. Our validation targets include a median geolocation error below 300 metres, an F1-score exceeding 0.80 for event-type classification, 90% taxonomy coverage of distinct high-impact events, and at least 70% of detected events providing positive lead time relative to official alerts. The paper concludes by discussing integration pathways through Mapbox-GL dashboards, Common Alerting Protocol feeds, open data publication, and capacity-building programmes that, together, translate analytical outputs into safer, more resilient urban futures.