Today, business decision-makers deal with increasingly complex problems that need to be solved by making accurate decisions in a short amount of time. The acceleration of decision-making processes is linked not only to growing global competitiveness but also to the need to quickly adapt to market changes and take advantage of emerging opportunities. In this context, the role of managers has become more complicated, requiring them not only to have sector-specific skills but also to be able to analyze and interpret large volumes of data and manage situations of uncertainty and variability. Given these complexities, the adoption of intelligent Decision Support Systems (DSS) has become essential for improving business performance. This thesis aims to study and implement some DSS that leverage advanced Artificial Intelligence and Operations Research solutions to address specific problems in different industrial contexts. Traditionally, managerial decisions are based on optimization problems, which can be represented as mathematical problems with well-defined objectives, such as minimizing costs or maximizing operational efficiency. These optimization problems can be solved using techniques that are widely studied in the Operations Research literature, such as mathematical models and heuristic algorithms. In recent years, the explosion of Big Data has radically transformed how companies collect and use information. Thanks to the large amount of data collected daily by companies, Artificial Intelligence techniques have also found widespread use in addressing optimization problems. Various techniques can be employed to make accurate predictions and evaluate future scenarios, allowing companies to anticipate changes and react quickly and efficiently. The outputs of these systems can be evaluated using specific indices and metrics that managers can use to improve the decision-making processes in various contexts. For this reason the integration of Artificial Intelligence and Operations Research techniques within DSS facilitates the effective resolution of specific challenges across diverse sectors, thereby enhancing the quality of decision-making processes. This thesis is structured into three main chapters, each featuring a real-world application of DSS: demand forecasting in the perishable food industry, online motion accuracy compensation of industrial servomechanisms, and energy-efficient scheduling. The work demonstrates how these technologies can be used not only to solve complex problems but also to make decision-making processes more flexible, responsive, and sustainable. By implementing these approaches in real-world contexts, the ultimate goal is to improve operational efficiency, reduce waste, and optimize the use of business resources, thus contributing to long-term value creation for companies.
Oggi, i decision-makers aziendali si trovano ad affrontare problemi sempre più complessi che richiedono decisioni accurate da prendere in tempi brevi. L’accelerazione dei processi decisionali è legata non solo alla crescente competitività globale, ma anche alla necessità di adattarsi rapidamente ai cambiamenti del mercato e di cogliere le opportunità emergenti. In questo contesto, il ruolo dei manager è diventato più complicato, richiedendo loro non solo competenze specifiche del settore, ma anche la capacità di analizzare e interpretare grandi volumi di dati e di gestire situazioni di incertezza e variabilità. Date queste complessità, l’adozione di Sistemi di Supporto alle Decisioni (DSS) intelligenti è diventata essenziale per migliorare le prestazioni aziendali. Questa tesi si propone di studiare e implementare alcuni DSS che sfruttano soluzioni avanzate di Intelligenza Artificiale e Ricerca Operativa per affrontare problemi specifici in diversi contesti industriali. Tradizionalmente, le decisioni manageriali si basano su problemi di ottimizzazione, che possono essere rappresentati come problemi matematici con obiettivi ben definiti, come minimizzare i costi o massimizzare l’efficienza operativa. Questi problemi di ottimizzazione possono essere risolti utilizzando tecniche ampiamente studiate nella letteratura della Ricerca Operativa, come modelli matematici e algoritmi euristici. Negli ultimi anni, l’esplosione dei Big Data ha trasformato radicalmente il modo in cui le aziende raccolgono e utilizzano le informazioni. Grazie alla grande quantità di dati raccolti quotidianamente dalle aziende, anche le tecniche di Intelligenza Artificiale hanno trovato largo impiego nella risoluzione di problemi di ottimizzazione. Varie tecniche possono essere impiegate per effettuare previsioni accurate e valutare scenari futuri, consentendo alle aziende di anticipare i cambiamenti e reagire in modo rapido ed efficiente. I risultati di questi sistemi possono essere valutati utilizzando indici e metriche specifiche che i manager possono sfruttare per migliorare i processi decisionali in diversi contesti. Per questo motivo l’integrazione delle tecniche di Intelligenza Artificiale e Ricerca Operativa all’interno dei DSS facilita la risoluzione efficace di sfide specifiche, migliorando così la qualità dei processi decisionali. Questa tesi è strutturata in tre capitoli principali, ciascuno caratterizzato da un’applicazione reale di DSS: previsione della domanda nell’industria alimentare deperibile, compensazione dell’errore di movimento online dei servomeccanismi industriali, ed energy-efficient scheduling. Questo dimostra come le tecnologie utilizzate possano essere utilizzate non solo per risolvere problemi complessi, ma anche per rendere i processi decisionali più flessibili, reattivi e sostenibili. Implementando questi approcci in contesti reali, l’obiettivo finale è migliorare l’efficienza operativa, ridurre gli sprechi e ottimizzare l’uso delle risorse aziendali, contribuendo così alla creazione di valore a lungo termine per le aziende.
Algoritmi di Ottimizzazione e Sistemi Intelligenti di Supporto alle Decisioni per l'industria di prossima generazione
MUCCIARINI, MIRKO
2025
Abstract
Today, business decision-makers deal with increasingly complex problems that need to be solved by making accurate decisions in a short amount of time. The acceleration of decision-making processes is linked not only to growing global competitiveness but also to the need to quickly adapt to market changes and take advantage of emerging opportunities. In this context, the role of managers has become more complicated, requiring them not only to have sector-specific skills but also to be able to analyze and interpret large volumes of data and manage situations of uncertainty and variability. Given these complexities, the adoption of intelligent Decision Support Systems (DSS) has become essential for improving business performance. This thesis aims to study and implement some DSS that leverage advanced Artificial Intelligence and Operations Research solutions to address specific problems in different industrial contexts. Traditionally, managerial decisions are based on optimization problems, which can be represented as mathematical problems with well-defined objectives, such as minimizing costs or maximizing operational efficiency. These optimization problems can be solved using techniques that are widely studied in the Operations Research literature, such as mathematical models and heuristic algorithms. In recent years, the explosion of Big Data has radically transformed how companies collect and use information. Thanks to the large amount of data collected daily by companies, Artificial Intelligence techniques have also found widespread use in addressing optimization problems. Various techniques can be employed to make accurate predictions and evaluate future scenarios, allowing companies to anticipate changes and react quickly and efficiently. The outputs of these systems can be evaluated using specific indices and metrics that managers can use to improve the decision-making processes in various contexts. For this reason the integration of Artificial Intelligence and Operations Research techniques within DSS facilitates the effective resolution of specific challenges across diverse sectors, thereby enhancing the quality of decision-making processes. This thesis is structured into three main chapters, each featuring a real-world application of DSS: demand forecasting in the perishable food industry, online motion accuracy compensation of industrial servomechanisms, and energy-efficient scheduling. The work demonstrates how these technologies can be used not only to solve complex problems but also to make decision-making processes more flexible, responsive, and sustainable. By implementing these approaches in real-world contexts, the ultimate goal is to improve operational efficiency, reduce waste, and optimize the use of business resources, thus contributing to long-term value creation for companies.File | Dimensione | Formato | |
---|---|---|---|
Tesi definitiva Mucciarini Mirko.pdf
embargo fino al 13/03/2026
Dimensione
13.77 MB
Formato
Adobe PDF
|
13.77 MB | Adobe PDF |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.14242/196044
URN:NBN:IT:UNIMORE-196044