Although traditional predictive methods support project planning, they often struggle to capture the high-dimensional interactions, nonlinearity, structural complexity, uncertainty and evolving nature of real-world projects. This limitation can reduce predictive accuracy, robustness and probabilistic reliability, thereby constraining managerial responsiveness. In response, this study is motivated by the limited number of field-validated empirical studies that apply advanced ensemble learning techniques, such as calibrated stacking, to model complex relationships among project variables and, in turn, predict project success. Accordingly, the objective of this study is to develop and rigorously evaluate a statistically validated ensemble learning framework for binary project success prediction under realistic, imbalanced project management conditions. The proposed framework integrates feature-selection enhancement, inferential statistical testing, multi-metric evaluation and embedded interpretability diagnostics. The results indicate that calibration-aware stacking with regularized meta-learning delivers robust predictive performance, offering more reliable analytics and stronger decision support for managers operating in complex project environments.
Grounded in middle-range theory (MRT), this study explores supply chain resilience (SCRes) during the 2021 Henan floods through 54 interviews with agri-food industry practitioners. The findings reveal SCRes to be a multi-level framework determined by individuals, organizations, supply chains, and broader environments collectively. The environmental level, comprising political, economic, social, technological, environmental, legal, and cultural (PESTELC) factors, functions as the resource layer, providing critical resources. The supply chain level is the direction layer, setting goals for preparation, response, recovery, and adaptation by the organizations and individuals involved. The organizational level is the transmission layer, disseminating resilience objectives both horizontally and vertically. Finally, the individual level is the implementation layer, operationalizing organizational and SCRes strategies. Organizational employees' personal attributes, such as openness, persistence, and extraversion, should be carefully considered when implementing SCRes strategies. China's hierarchical culture, chain leader system, and accountability mechanisms ensure coordination across all levels.
This paper examines the current state of research, key aspects, and implications of the relationships between Industry 4.0 technologies (I4.0T), lean management practices (LMPs), and sustainable operational performance (SOPR). Using a theory-driven lens, the study conducts a systematic literature review (SLR) of 50 peer-reviewed articles to explore synergies, tensions, and research gaps within this emerging domain. Drawing on Socio-Technical Systems (STS) and Complementarity Theory, the review identifies a core paradox: while digitalisation enhances lean performance through improved visibility, connectivity, and process integration, it also creates human-centric tensions related to autonomy, identity, learning demands, workforce adaptation, and organisational flexibility. The findings show that synergistic integration of LMPs and I4.0T can improve operational, social, and environmental outcomes, whereas misalignments between automation and human-centred practices generate paradoxical tensions. The paper proposes a future research agenda focused on paradox management, adaptability, and transitions towards Industry 5.0 systems.
Grounded in social exchange and social network theories, we study whether and how buyer firms' social sustainability draws on their suppliers' digital technology-oriented open innovation (DTOI). We focus on two dimensions of buyer social sustainability, namely societal and workforce sustainability, and examine the moderating role of buyers' director network structure. Using panel data on Chinese listed firms from 2014 to 2024, we find that both dimensions are positively associated with supplier DTOI. These relationships are stronger for buyers whose director networks feature higher degree centrality, betweenness centrality, and structural holes, but not closeness centrality. We further explore how these relationships vary across relational, ownership, and industry contexts. Our findings extend digital innovation research to supply chains and clarify when supplier spillovers reach buyers.
Generative AI (GenAI) is a promising enabler of predictive maintenance in capital-intensive manufacturing. However, there is limited empirical evidence on how it can be effectively embedded in production planning and control and on the performance gains it delivers. This paper reports an in-depth case study of ABC Tyres Limited (ABC), a tyre manufacturer that introduced a GenAI-enabled predictive maintenance solution for its curing presses. The solution combines a domain-adapted large language model (LLM) with the plant's process data lake, allowing engineers to query historical production cycles in natural language and generate diagnostic insights. Using before-and-after operational data, we assess the impact on key performance indicators. The intervention reduced the mean dry cycle time from 106 to 80 seconds, stabilized the curing process, increased daily curing capacity by 304 tyres, reduced steam consumption by 0.3 MT per day, and shortened the turnaround time for complex root cause analyses from around 10 days to roughly 2 hours.
