Recently Industry 4.0 was introduced and with the help of its key enabling technologies started to revolutionize human work creating intelligent manufacturing systems. The main goals of Industry 4.0 are to achieve a higher level of operational efficiency and productivity, as well as a higher level of automatization while maximizing productivity. For reaching the said goals the key enabling technologies are Artificial Intelligence (AI), Additive Manufacturing (AM), IoT systems, Big Data, reinforcement learning and cloud computing. In this thesis starting from current literature the application of these technologies for the management of Industrial Plants has been investigated. In particular, regarding operational technologies the use of AM for the management of spare parts has been investigated given the challenges for a correct spare parts management hindered by Recently Industry 4.0 was introduced and with the help of its key enabling technologies started to revolutionize human work creating intelligent manufacturing systems. The main goals of Industry 4.0 are to achieve a higher level of operational efficiency and productivity, as well as a higher level of automatization while maximizing productivity. For reaching the said goals the key enabling technologies are Artificial Intelligence (AI), Additive Manufacturing (AM), IoT systems, Big Data, reinforcement learning and cloud computing. In this thesis starting from current literature the application of these technologies for the management of Industrial Plants has been investigated. In particular, regarding operational technologies the use of AM for the management of spare parts has been investigated given the challenges for a correct spare parts management hindered by their intermittent demand and high cost of shortage. The situations that favor the insourcing of AM have been defined in various contexts: spare parts on demand production, insourced AM as a resilience tool for spare parts supply chain and AM for spare parts preventive maintenance. On the same line of research regarding operational technologies, the role of low-cost sensors for the evaluation of the integrated risk to which operators are subjected has been investigated in combination with an indirect eliciting of risk category weights. Indirect eliciting that has been compared with classical direct eliciting finding the superiority of an indirect eliciting in reconstructing operators’ preferences. This stream of research started from the need to consider humans in the loop that were ignored by Industry 4.0 and thus refer to the newly started Industry 5.0. Low-cost sensors have been investigated also in the light of ergonomic risk developing two different applications based on the new depth camera Azure Kinect to evaluate it semi-automatically. Lastly, in the light of informational technologies two different technologies have been tested: Big Data analytics have been exploited to facilitate the management of spare parts and dynamic programming for the re-order of items in high-volatile markets with demand updates.
Recentemente l’Industria 4.0 ha rivoluzionato il mondo del lavoro creando sistemi di produzione intelligenti grazie alle sue tecnologie abilitanti. Gli obiettivi principali dell'Industria 4.0 sono il raggiungimento di un più alto livello di efficienza operativa e produttiva, produttività aumentata grazie all’automazione dei processi. Per raggiungere tali obiettivi, le tecnologie chiave abilitanti dell’Industria 4.0 sono: Intelligenza Artificiale (IA), Additive Manufacturing (AM), sistemi IoT, Big Data, reinforcement learning e cloud computing. In questa tesi, partendo dallo studio della letteratura attuale, è stata valutata l'applicazione di queste tecnologie per la gestione di impianti industriali. In particolare, per quanto riguarda le tecnologie operative, è stato esaminato l’utilizzo di AM per la gestione dei pezzi di ricambio. Questo in quanto la gestione delle parti di ricambio è ostacolata dalla loro domanda intermittente e dagli alti costi di fermo impianto derivanti dall’assenza di una parte di ricambio. Sono state ricavate sperimentalmente le situazioni che favoriscono l’adozione di AM in vari contesti: l’insourcing di AM per la produzione in tempo reale delle parti di ricambio, l’insourcing di AM come strumento per aumentare la resilienza della catena di fornitura di parti di ricambio e AM per la manutenzione preventiva delle parti di ricambio. Sempre riguardo le tecnologie operative è stata esaminato il ruolo di sensori a basso costo per la valutazione del rischio integrato a cui sono soggetti gli operatori in ambito industriale in combinazione con un eliciting indiretto delle categorie di rischio. L'eliciting indiretto è stato confrontato con il più classico eliciting diretto. Confronto che ha evidenziato la superiorità di un eliciting indiretto nel ricostruire le preferenze degli operatori. Questo filone di ricerca è nato dalla necessità di includere gli operatori e le loro caratteristiche univoche nella gestione degli impianti industriali dando importanza al fattore umano che era stato trascurato dall’industria 4.0 e che diventa ora un pilastro della nuova Industria 5.0. Sensori a basso costo per la valutazione del rischio sono stati esaminati anche alla luce del calcolo del rischio ergonomico, sviluppando due diverse applicazioni basate sulla nuova telecamera di profondità Azure Kinect. Infine, alla luce delle tecnologie dell'informazione, sono state testate due diverse tecnologie: l'analisi dei Big Data attraverso algoritmi di clustering è stata sfruttata per facilitare il riordino combinato dei pezzi di ricambio mentre la programmazione dinamica è stata sfruttata per il riordino di articoli in mercati ad alta volatilità.
Gestione delle scorte nell'era dell'industria 4.0.
CORUZZOLO, ANTONIO MARIA
2024
Abstract
Recently Industry 4.0 was introduced and with the help of its key enabling technologies started to revolutionize human work creating intelligent manufacturing systems. The main goals of Industry 4.0 are to achieve a higher level of operational efficiency and productivity, as well as a higher level of automatization while maximizing productivity. For reaching the said goals the key enabling technologies are Artificial Intelligence (AI), Additive Manufacturing (AM), IoT systems, Big Data, reinforcement learning and cloud computing. In this thesis starting from current literature the application of these technologies for the management of Industrial Plants has been investigated. In particular, regarding operational technologies the use of AM for the management of spare parts has been investigated given the challenges for a correct spare parts management hindered by Recently Industry 4.0 was introduced and with the help of its key enabling technologies started to revolutionize human work creating intelligent manufacturing systems. The main goals of Industry 4.0 are to achieve a higher level of operational efficiency and productivity, as well as a higher level of automatization while maximizing productivity. For reaching the said goals the key enabling technologies are Artificial Intelligence (AI), Additive Manufacturing (AM), IoT systems, Big Data, reinforcement learning and cloud computing. In this thesis starting from current literature the application of these technologies for the management of Industrial Plants has been investigated. In particular, regarding operational technologies the use of AM for the management of spare parts has been investigated given the challenges for a correct spare parts management hindered by their intermittent demand and high cost of shortage. The situations that favor the insourcing of AM have been defined in various contexts: spare parts on demand production, insourced AM as a resilience tool for spare parts supply chain and AM for spare parts preventive maintenance. On the same line of research regarding operational technologies, the role of low-cost sensors for the evaluation of the integrated risk to which operators are subjected has been investigated in combination with an indirect eliciting of risk category weights. Indirect eliciting that has been compared with classical direct eliciting finding the superiority of an indirect eliciting in reconstructing operators’ preferences. This stream of research started from the need to consider humans in the loop that were ignored by Industry 4.0 and thus refer to the newly started Industry 5.0. Low-cost sensors have been investigated also in the light of ergonomic risk developing two different applications based on the new depth camera Azure Kinect to evaluate it semi-automatically. Lastly, in the light of informational technologies two different technologies have been tested: Big Data analytics have been exploited to facilitate the management of spare parts and dynamic programming for the re-order of items in high-volatile markets with demand updates.File | Dimensione | Formato | |
---|---|---|---|
Tesi_PhD_AMC_REV_fin.pdf
embargo fino al 20/03/2027
Dimensione
4.27 MB
Formato
Adobe PDF
|
4.27 MB | Adobe PDF |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.14242/80207
URN:NBN:IT:UNIMORE-80207