In volatile manufacturing environments, particularly in project-based plant engineering, servitization is regarded as a strategic approach to enhance competitiveness and resilience. However, empirical evidence on the relationship between servitization capabilities and firm- and/or network-level performance remains limited. This study develops and tests a multi-level framework to examine how classical and digital servitization capabilities, enabled by predictive/analytical artificial intelligence (AI), link to operational performance. A sequential mixed-methods design combined expert interviews and structural equation modelling based on survey data. Qualitative findings identified digitalization, customer relationship management, and operational excellence as key capability domains. Quantitative results show that neither classical nor digital servitization capabilities significantly link to firm-level performance. Instead, both capabilities strongly link to network-level coordination, collective operational performance, and supply chain resilience. The findings shift the performance focus of servitization from individual firms to manufacturing ecosystems and highlight network orchestration and capability renewal as key drivers of competitive advantage.
This paper introduces an index that directly measures how much of a buyer’s existing supplier base is replaced from one period to the next. We construct this “fraction of supplier share dropped” measure using data on 229 automotive parts, 693 suppliers, and 12 Japanese automakers from 1996 to 2016. Our findings indicate that Toyota, known for stable ties with its affiliates, exhibits the lowest switching rate among those in-group suppliers; however, it shows one of the highest switching rates when the suppliers are not within its affiliate network. Other automakers exhibit less dramatic shifts in switching behavior across affiliated and non-affiliated suppliers. By capturing the identity-specific nature of supplier turnover, our index advances the literature on buyer–supplier dynamics and offers managers a more precise tool for assessing the trade-offs between stable, in-group relationships and the strategic benefits of reevaluating out-group suppliers.
The Digital Twin (DT) plays a crucial role in the digital transformation era by providing a bidirectional synchronized digital representation of an asset, capable of performing real-time monitoring, simulation, and data analysis. Generally, DTs are designed as single units following a centralised approach. However, this paradigm is evolving towards DT ecosystems, consisting of a network of multiple interconnected DTs. The benefits of such ecosystems depend on their intended purpose and application, which are strongly influenced by the design and structural configuration adopted during implementation. This paper aims to analyse different organisational structures, namely centralised, hierarchical, heterarchical, and holonic, for the design, development, and implementation of DT ecosystems, assessing their benefits and challenges across a set of aspects. A case study of a modular conveyor transfer system was used to test the different organisational structures, supporting the comparative analysis and the discussion of the research challenges related to this topic.
The role of the maintenance operator is evolving with Industry 4.0, leveraging advanced technologies to enhance physical, sensory, and cognitive capabilities. Efficient troubleshooting support is critical to reducing operators’ physical and mental stress in high-demand maintenance environments. Augmented reality (AR) presents a powerful solution by integrating real-time data visualisation, contextualised fault diagnostics, and interactive guidance into the troubleshooting workflow. However, existing research lacks structured approaches for AR-assisted troubleshooting that enable operators to rapidly detect, analyse, and resolve system failures. This research investigates the essential artificial intelligence (AI) and AR functionalities needed for effective troubleshooting. It examines how AR can enhance troubleshooting by seamlessly integrating real-time data, adaptive user interfaces (UI), and operator-driven insights. Findings from a railway case study indicate that AR-assisted troubleshooting improves decision-making, reduces task errors, facilitates knowledge capture, and supports both novice and experienced operators in structuring and contextualising system data. Iterative prototype development highlighted the importance of clear visualisation, guided autonomy, contextual information, and operator feedback. Future research will explore adaptive interfaces, real-time feedback loops, and broader deployment to further optimise AR-supported maintenance workflows.
Production environments and their constituent assets often deviate from their original design intent because of design assumptions, variability in materials and processes, and external stressors. Furthermore, assets are not always used as intended, requiring mechanisms to validate their performance over time. Traditionally, commissioning ensures compliance with client requirements before operational deployment but is typically treated as a one-time event. This study extends the concept of commissioning to support ongoing validation throughout the entire asset lifecycle using a commissioning-based framework. This framework enables the assessment of current configurations and exploration of future scenarios, allowing for context-aware (re)commissioning by balancing a priori and a posteriori activities. Commissioning thereby shifts to an integral and iterative process that facilitates alignment between evolving requirements and actual system behaviour, supporting better-informed decision-making.
