A FRAMEWORK TO GUIDE SUSTAINABLE STRATEGIC, TACTICAL AND OPERATIONAL DECISION MAKING IN MANUFACTURING ORGANIZATIONS

The transition of the manufacturing sector toward CE models and industrial sustainability requires overcoming the traditional linear paradigm and integrating the enabling technologies of Industry 4.0. However, the scientific literature reveals significant theoretical and methodological fragmentation, characterized by a proliferation of DSS applied primarily to isolated case studies and lacking generalizable selection principles capable of linking corporate strategic directives to shop floor physical control. To bridge this gap, this thesis develops a unified conceptual framework and a five-step decision characterization protocol designed to guide the rigorous selection of algorithmic architectures based on the intrinsic properties of the industrial problem. The methodology relies on a systematic literature review of 45 foundational scientific contributions indexed in the Scopus database, with extracted data organized into cross-tabulation matrices and decision trees across four management horizons: strategic, tactical, operational, and multilevel. Empirical analysis formalizes the dominant rules and methodological pathways governing problem resolution across organizational echelons. At the strategic level, evaluating discrete alternatives under deterministic or fuzzy conditions favors multicriteria decision analysis to formulate preference rankings and recommended strategies supporting long-term investments and product end of life options. At the tactical level, midterm planning and balancing material flows against production capacities under stochastic uncertainty converge on mathematical programming and hybrid optimization models for resource allocation. At the operational level, machine parameter tuning uses hybrid metaheuristics and response surface models to calculate exact numerical values for direct input into machine tool controllers. Unlike single horizon levels, which are characterized by monolithic decision flows traversing predefined structural steps, the multilevel domain does not follow a single rigid pathway; decision pathways at this level are instead constructed step by step, combining the fundamental patterns based on cross variable interactions identified within the taxonomy. Concurrently, the research demonstrates a systematic alignment between decision objects and adopted sustainability frameworks, wherein physical plant and process operations favor bi-dimensional economic and environmental metrics, decisions regarding products and logistics networks require comprehensive evaluations based on the three pillars of the TBL, and corporate entities map onto ESG reporting frameworks to satisfy regulatory transparency mandates and investor benchmark compliance. The contributions of this thesis provide a prescriptive methodological guide to identify the most suitable decision support tool based on specific sustainability objectives and problem characteristics