In this thesis work, an ontology was built to solve problems related to the Mechanical Engineering world. In the first chapters, an overview with some examples was made, related to Artificial Intelligence, Semantic WEB, ontologies and knowledge graph population. This analysis of the main characteristics was done using the knowledge shared by Professor Harald Sack. In particular, ontologies will be described in detail, analysing their structure and different ways in which they can be employed in Manufacturing. The pairing of both Artificial Intelligence and ontologies with Digital Twin will also be briefly analysed. Then, a problem coming from real-life experiences of mechanical engineers will be presented, concerning milling: in particular, it regards the choice of the most suitable milling tool for machining a certain product. The selection process without ontologies is presented and all the constraints are underlined. Then, an ontology is developed using Protégé, presenting the various stages of development with a "sprint" format. The choice of the ontology is related to its intrinsic capacity of representing hierarchies and properties, which applies very well to the representation of a milling tool catalogue. The final result will follow the LOT procedure. This last result will be validated in different ways, highlighting its strengths and possible improvements. Finally, developments that could be pursued in the future will be enlisted.
Per questo lavoro di tesi, un’ontologia è stata costruita per risolvere problemi legati al mondo dell’Ingegneria Meccanica. Nei primi capitoli, una panoramica completa di esempi è stata fatta, per argomenti come Intelligenza Artificiale, Semantic WEB, ontologie e come popolare un knowledge graph. L’analisi delle principali caratteristiche è stata fatta utilizzando la conoscenza messa a disposizione dal Professor Harald Sack. In particolare, le ontologie verrano descritte nel dettaglio, analizzando la loro struttura e in che modo possono essere impiegate nel mondo del Manufacturing. L’abbinamento sia di Intelligenza Artificiale che di ontologie con il concetto di Digital Twin verrà brevemente analizzato. Poi, un problema da esperienze reali di ingegneri meccanici verrà presentato, riguardante la fresatura: in particolare, riguarda le scelta dell’utensile di frsatura più adeguato per produrre un certo pezzo. Il processo di selezione senza l’uso di ontologie viene presentato e tutti i vincoli vengono sottolineati. Successivamente, un’ontologia verrà sviluppata usando Protégé [31], presentando i diversi stadi di sviluppo con un formato a "sprint". La scelta di sviluppare un’ontologia è legata alla sua intrinseca capacità di rappresentare le gerarchie e le proprietà, caso che si applica molto bene alla rappresentazione di un catalogo di utensili da fresatura. In risultato finale seguirà la procedura LOT. Quest’ultimo risultato verrà validato in diversi modi, evidenziando i suoi punti di forza e possibili miglioramenti. Infine, possibili sviluppi per il futuro verranno elencati.
Knowledge graphs to support milling tool selection
SALLUZZI, GIUSEPPINA
2023/2024
Abstract
In this thesis work, an ontology was built to solve problems related to the Mechanical Engineering world. In the first chapters, an overview with some examples was made, related to Artificial Intelligence, Semantic WEB, ontologies and knowledge graph population. This analysis of the main characteristics was done using the knowledge shared by Professor Harald Sack. In particular, ontologies will be described in detail, analysing their structure and different ways in which they can be employed in Manufacturing. The pairing of both Artificial Intelligence and ontologies with Digital Twin will also be briefly analysed. Then, a problem coming from real-life experiences of mechanical engineers will be presented, concerning milling: in particular, it regards the choice of the most suitable milling tool for machining a certain product. The selection process without ontologies is presented and all the constraints are underlined. Then, an ontology is developed using Protégé, presenting the various stages of development with a "sprint" format. The choice of the ontology is related to its intrinsic capacity of representing hierarchies and properties, which applies very well to the representation of a milling tool catalogue. The final result will follow the LOT procedure. This last result will be validated in different ways, highlighting its strengths and possible improvements. Finally, developments that could be pursued in the future will be enlisted.File | Dimensione | Formato | |
---|---|---|---|
2024_12_Salluzzi.pdf
accessibile in internet solo dagli utenti autorizzati
Descrizione: thesis text
Dimensione
7.58 MB
Formato
Adobe PDF
|
7.58 MB | Adobe PDF | Visualizza/Apri |
2024_12_Salluzzii_executivesummary.pdf
accessibile in internet per tutti
Descrizione: executive summary
Dimensione
1.04 MB
Formato
Adobe PDF
|
1.04 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/230024