The work addresses the problem of traceability and process control in robotic lines dedicated to the assembly of composite components or sub-assemblies of the bodywork for autonomous vehicles. In particular, the complexity of composite materials (sheet molding compound) and surface preparation, combined with the sensitivity of two-component structural adhesives, requires a system capable of collecting, validating, and correlating critical process data. To this end, the study focused on defining a methodology, or engineering approach, to assist in the implementation and validation of a Manufacturing Execution System (MES) management system. The methodology adopted consists of four main phases. The first involves the preliminary analysis of the data available on the production line, through a pre-automation phase in which the consistency, stability, and reliability of the data from the Programmable Logic Controllers (PLCs) was verified before validating it as useful information. Subsequently, the MES-PLC interface architecture was defined, which made it possible to structure the information flow necessary for the traceability of individual components. On this basis, the third phase was developed, with the design of the data model and the selection of Key Performance Indicators (KPIs) relevant to the assembly department. The final phase involved assessing process capacity and actively monitoring information using dashboards and key indices, such as Overall Equipment Effectiveness (OEE), which improved early diagnosis capabilities and visibility of line performance. Possible traceability implementations, such as outlet pressure, were proposed. The analysis of the results was then presented, with particular attention to the critical issues that emerged along the production flow, from which improvement proposals and potential future implementations were derived. The results show that the system, which communicates with the robotic lines and the company's Enterprise Resource Planning (ERP) management system, allows each car body to be associated with the serial numbers of the sub-assemblies, such as the side panels or roof, made of SMC composite materials, and with the actual values of the joining process of the various parts first and then the various sub-assemblies of the body, through the use of dedicated lines. Data validation has highlighted the possibility of reconstructing the complete genealogy of the component and identifying correlations related to deviations in the catalysis ratio. In addition to enabling automatic warehouse picking and depositing, integration has also made it possible to monitor process stability in real time, reducing variability due to both material and operating conditions. The conclusions confirm that the introduction of MES in the gluing department is a key factor in joint quality and process repeatability, transforming a highly sensitive production sequence into a controlled and traceable system. The main limitations concern dependence on the quality of PLC signals and the need to completely standardize data records and data flows. It is recommended to extend data collection through further implementations, such as the involvement of operators or the inclusion of dedicated alarms, and to integrate predictive algorithms based on bonding parameters, so as to consolidate a data-driven maintenance model to further increase process reliability.
Il lavoro affronta il problema della tracciabilità e del controllo di processo nelle linee robotizzate dedicate all’assemblaggio di componenti, o sottogruppi della scocca, in materiale composito destinati a veicoli a guida autonoma. In particolare, la complessità dei materiali compositi: Sheet Moulding Compound e della preparazione superficiale, unita alla sensibilità degli adesivi strutturali bicomponenti, rende necessario un sistema capace di raccogliere, validare e correlare i dati critici del processo. A questo fine, lo studio si è focalizzato sulla definizione di una metodologia, o approccio ingegneristico, che aiuti nell'implementazione e validazione di un sistema gestionale di tipo Manufacturing Execution System (MES). La metodologia adottata si articola in quattro fasi principali. La prima riguarda l’analisi preliminare dei dati disponibili a bordo linea, attraverso una fase di preautomazione in cui è stata verificata la coerenza, la stabilità e l’affidabilità dei dati provenienti dai Programmable Logic Controller (PLC), prima di validarli come informazioni utili. Successivamente, è stata definita l’architettura di interfacciamento MES–PLC, che ha permesso di strutturare il flusso informativo necessario alla tracciabilità del singolo componente. Su questa base si è sviluppata la terza fase, con la progettazione del modello dati e la selezione dei Key Performance Indicator (KPI) rilevanti per il reparto di assemblaggio. L’ultima fase ha riguardato la valutazione della capacità di processo e il monitoraggio attivo delle informazioni tramite l’utilizzo di dashboard e indici principali, come l’Efficacia Complessiva dell’Impianto (OEE), hanno migliorato la capacità di diagnosi precoce e la visibilità sulle prestazioni della linea. Possibili implementazioni lato tracciabilità, come la pressione in uscita, sono state proposte. È stata quindi presentata l’analisi dei risultati, con particolare attenzione alle criticità emerse lungo il flusso produttivo, da cui sono state derivate proposte migliorative e potenziali implementazioni future. I risultati mostrano che il sistema, comunicante con le linee robotizzate e con il gestionale aziendale di Pianificazione delle Risorse d'Impresa (ERP), consente di associare ogni scocca vettura ai seriali dei sottogruppi, come fiancata o tetto, realizzati in materiali compositi SMC e ai valori effettivi del processo di giunzione delle varie parti prima e dei vari sottogruppi della scocca dopo, tramite l’utilizzo delle linee dedicate. La validazione dei dati ha evidenziato la possibilità di ricostruire la genealogia completa del componente e di individuare correlazioni legate alle deviazioni del rapporto di catalisi. L’integrazione, oltre ad aver reso possibile la gestione di prelievo e versamento automatico del magazzino, ha inoltre permesso di monitorare in tempo reale la stabilità del processo, riducendo la variabilità dovuta sia al materiale sia alle condizioni operative. Le conclusioni confermano che l’introduzione del MES nel reparto di incollaggio rappresenta un elemento abilitante per la qualità del giunto e per la ripetibilità del processo, trasformando una sequenza produttiva altamente sensibile in un sistema controllato e tracciabile. Le principali limitazioni riguardano la dipendenza dalla qualità dei segnali PLC e la necessità di standardizzare completamente le anagrafiche e i flussi dati. Si raccomanda di estendere la raccolta dati tramite ulteriori implementazioni, come il coinvolgimento degli operatori o l’inserimento di allarmi dedicati, e di integrare algoritmi predittivi basati sui parametri di incollaggio, così da poter consolidare un modello di manutenzione data‑driven per aumentare ulteriormente l’affidabilità del processo.
IMPLENTAZIONE E VALIDAZIONE DI UN SISTEMA MES SU LINEE ROBOTIZZATE DI ASSEMBLAGGIO VEICOLI
GIOVANE, MASSIMO
2024/2025
Abstract
The work addresses the problem of traceability and process control in robotic lines dedicated to the assembly of composite components or sub-assemblies of the bodywork for autonomous vehicles. In particular, the complexity of composite materials (sheet molding compound) and surface preparation, combined with the sensitivity of two-component structural adhesives, requires a system capable of collecting, validating, and correlating critical process data. To this end, the study focused on defining a methodology, or engineering approach, to assist in the implementation and validation of a Manufacturing Execution System (MES) management system. The methodology adopted consists of four main phases. The first involves the preliminary analysis of the data available on the production line, through a pre-automation phase in which the consistency, stability, and reliability of the data from the Programmable Logic Controllers (PLCs) was verified before validating it as useful information. Subsequently, the MES-PLC interface architecture was defined, which made it possible to structure the information flow necessary for the traceability of individual components. On this basis, the third phase was developed, with the design of the data model and the selection of Key Performance Indicators (KPIs) relevant to the assembly department. The final phase involved assessing process capacity and actively monitoring information using dashboards and key indices, such as Overall Equipment Effectiveness (OEE), which improved early diagnosis capabilities and visibility of line performance. Possible traceability implementations, such as outlet pressure, were proposed. The analysis of the results was then presented, with particular attention to the critical issues that emerged along the production flow, from which improvement proposals and potential future implementations were derived. The results show that the system, which communicates with the robotic lines and the company's Enterprise Resource Planning (ERP) management system, allows each car body to be associated with the serial numbers of the sub-assemblies, such as the side panels or roof, made of SMC composite materials, and with the actual values of the joining process of the various parts first and then the various sub-assemblies of the body, through the use of dedicated lines. Data validation has highlighted the possibility of reconstructing the complete genealogy of the component and identifying correlations related to deviations in the catalysis ratio. In addition to enabling automatic warehouse picking and depositing, integration has also made it possible to monitor process stability in real time, reducing variability due to both material and operating conditions. The conclusions confirm that the introduction of MES in the gluing department is a key factor in joint quality and process repeatability, transforming a highly sensitive production sequence into a controlled and traceable system. The main limitations concern dependence on the quality of PLC signals and the need to completely standardize data records and data flows. It is recommended to extend data collection through further implementations, such as the involvement of operators or the inclusion of dedicated alarms, and to integrate predictive algorithms based on bonding parameters, so as to consolidate a data-driven maintenance model to further increase process reliability.| File | Dimensione | Formato | |
|---|---|---|---|
|
Tesi_Magistrale_GIOVANE_27_03_.pdf
embargo fino al 07/04/2029
Dimensione
7.26 MB
Formato
Adobe PDF
|
7.26 MB | Adobe PDF |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.14251/5255