Knowledge management (KM) has been gradually established as a strong methodology to support business viability, competitiveness and growth; however, the lack of maturity is obvious, as evidenced by divergent points of view in critical sub‐domains of the related theory and practice, such as the spectrum of processes and the clusters of tactics scheduled to underpin them. The systems thinking logic, the discipline for seeing structures underlying complex phenomena, is perceived as a promising direction for consolidating the various approaches and developing a holistic KM paradigm. Through the integration of the fragmented landscape of knowledge management on a transparent and concrete framework based on cause‐effect relationships, not only the underlying theoretical assumptions are clarified and subjected to further analysis at a strategic level, but also practical issues concerning planning and decision making become less complicated, increasing effectiveness.

In this work, two novel methodologies with the ability to address dependence and feedback among criteria and their interrelations are proposed for Cognitive Multi-Criteria Decision Making (CMCDM). Both methods, a modified Fuzzy Digraph approach and an FCM-based approach are intended for application in industrial robot selection. FCMs offer an alternative research direction in decision support that extends their traditional applications. Both methods are capable of handling crisp and fuzzy data; nevertheless, the FCM-based method offers a distinct advantage by enabling the evaluation of different scenarios based on the causal relationships between criteria. Moreover, a modification of Spearman’s rank correlation coefficient is presented, allowing for more meaningful comparisons between MCDM methods. Finally, an experimental analysis evaluates the effectiveness of the proposed methodologies, demonstrating their applicability in the field.

Shutdown, Turnaround, and Outage (STO) projects dealing with preventive, proactive or reactive maintenance are among the most complex and challenging ones in process industries. Their management is improved by incorporating advances in digital technologies and, particularly, in data analytics, computational and artificial intelligence (AI), leading to “Data-Driven Project Management (DDPM)”. A key issue arising is the type/nature of data that is used to support DDPM. We herein identify the key attributes of such data and show that they form a multidimensional set. Further, we explain how this multidimensional nature shapes the approaches to capture and analyze data and highlight the interplay between data analytics and the digital transformation that companies undergo, to increase their PM maturity. We demonstrate how multidimensional data analysis (MDDA) and information visualization (IV) leverage PM efficiency, and inversely, how the increased need for holistic PM triggers the development of sophisticated data analysis algorithms and tools. We critically evaluate how and to what extent, the advent of machine learning (ML) can supplement/replace existing approaches and highlight experiences from recent, pragmatic turnaround projects. A set of implemented “digital best practices” corroborates the importance of MDDA and IV in this family of projects and favors their use.