Production layout planning is under increasing pressure to adapt, as volatile demand, high product variety, and short product life cycles are pushing static layout concepts to their limits. This challenge is evident in control cabinet manufacturing, which is characterized by customer-specific products, batch size one, and production processes that remain strongly manual. Although reconfigurable manufacturing systems offer a promising solution, a coherent and application-oriented planning logic for reconfigurable robotic platforms at production-cell level is still lacking. Against this background, the paper uses a systematic literature review to examine which layout-related requirements and interface requirements for reconfigurable manufacturing systems (RMS)-based robotic platforms are reported in the literature and how these can be synthesized into a conceptual planning model for control cabinet manufacturing. The findings show that layout planning must jointly consider product-family fit, anticipatory adaptability, granular scalability, spatial configuration, low-effort reconfigurability, diagnosability, and safety-oriented observability. In parallel, the integration of reconfigurable robotic platforms requires a cross-level interface architecture spanning system, cell, logistics, and human levels, including standardized mechanical, electrical, and software interfaces, explicit factory-integration scenarios, consistent transfer and workpiece interfaces, and coordinated communication and data structures. The paper’s main contribution is the synthesis of these dispersed requirements into a five-step conceptual planning model, further specified through eight operational design rules for the production layout planning of reconfigurable robotic platforms. In doing so, the study closes the gap between generic RMS principles and concrete robotic-cell layout decisions and provides a conceptually grounded basis for incremental, spatially and organizationally integrated automation in control cabinet manufacturing.

This paper presents a comprehensive methodology for constructing a digital twin of a human operator to support ergonomic assessment in human–robot interaction (HRI) scenarios. Building on recent advances in computer vision and artificial intelligence, the proposed framework integrates video acquisition, 2D pose estimation, monocular depth inference, and 3D reconstruction to generate a temporally consistent representation of the operator. This digital twin enables automated evaluation of posture quality by computing anatomically relevant joint angles, including trunk flexion, neck inclination, shoulder elevation, elbow flexion, and knee bending.

To complement posture monitoring, the system incorporates an online fatigue estimation module that infers fatigue risk from cumulative exposure to non-neutral postures and repetitive movements. By comparing joint-angle trajectories against ergonomic thresholds, the method identifies potentially harmful configurations and provides real-time visual feedback to encourage posture correction. The proposed approach operates non-intrusively, requiring only a single monocular camera, and has been validated in an industrial pilot line involving continuous collaborative tasks.

Experimental results demonstrate that integrating real-time ergonomic recommendations slows fatigue accumulation and promotes healthier posture, thereby improving operator well-being. Overall, this work provides a scalable and sensor-free solution for proactive ergonomic monitoring in industrial HRI environments.

Feature-based feedback in Computer-Aided Design (CAD) for metal Additive Manufacturing (AM) offers an intelligent and automated way to assess geometric manufacturability in real time by identifying critical parametric features such as thin walls, overhangs, hole diameters, and unsupported spans. By comparing these features with machine-specific constraints, material behaviour and process physics. The system can provide immediate guidance on whether a design is suitable for metal AM. This helps designers minimize support structures, optimize build orientation, and reduce defect-prone geometries, which in turn improves build efficiency, material utilization, and overall part quality. When integrated with data-driven models and feature-extraction algorithms tailored to laser powder bed fusion (LPBF) and related AM processes. The approach further strengthens the link between design intent and manufacturing capability. Overall, this feedback mechanism supports the more reliable, sustainable and performance-optimized development of metal AM components.

The digitalization of production is considered a key driver of future manufacturing and thus a strategic goal. Despite notable advances in digital twins, innovative engineering, and individualized production, the expected dynamics of innovation and digital value creation have not materialized on a large scale, especially for small and medium-sized enterprises (SMEs). This paper argues that the root cause is structural. While highly digitalized industries are software-defined and based on ecosystems, the manufacturing sector continues to operate utilizing isolated, monolithic systems. An analysis of existing ecosystem approaches reveals that software-defined manufacturing cannot be based on a single ecosystem, but rather on multiple relevant types of ecosystems. Therefore, an ecosystem of ecosystems is suggested as the foundation for software-defined manufacturing. Furthermore, it is concluded that ecosystems cannot be developed by traditional means. Rather, they must be the result of an iterative process between their capabilities and participants. An analysis of key enablers reveals that they are shared by multiple ecosystems. Six cross-ecosystem key enablers are identified. Each is analyzed to determine its current state and deficits. Additionally, the potentialities that emerge as the key enabler advances are outlined. Steps and measures are suggested to evolve the key enablers and contribute to the formation of an ecosystem of ecosystems.

Supply chain disruptions caused by supplier failures and transportation interruptions increasingly require rapid operational adjustments rather than full network redesign. In many real-world settings, decision makers must modify an existing baseline plan under contractual and capacity constraints, while attempting to limit performance degradation. This paper develops a bi-objective mathematical model to support such operational reconfiguration decisions. The model preserves the pre-disruption configuration and enables corrective actions through supplier reallocation, transportation mode switching, and activation of Manufacturing-as-a-Service suppliers. Two conflicting objectives are considered: minimizing total reconfiguration cost and minimizing weighted unmet component demand as a direct measure of resiliency. We solve the resulting multi-objective problem by using the Augmented Epsilon-Constraint method to generate Pareto-efficient reconfiguration plans. Computational experiments based on disruption scenarios with varying structural characteristics reveal clear trade-offs among cost, service performance, and the degree of structural change. To enable decision making, we introduce knee-point identification and a structural reconfiguration intensity metric that quantifies deviation from the initial configuration. The results show that moderate structural adjustments can significantly improve service levels without imposing disproportionate cost increases.