Emaan Khan
Optimising Retrieval-Augmented Generation for Industrial Collaborative Robot Programming.
Rel. Giulia Bruno, Enrico Botticella. Politecnico di Torino, Corso di laurea magistrale in Ingegneria Gestionale (Engineering And Management), 2026
|
Preview |
PDF (Tesi_di_laurea)
- Tesi
Licenza: Creative Commons Attribution Non-commercial No Derivatives. Download (4MB) | Preview |
|
|
Archive (ZIP) (Documenti_allegati)
- Altro
Licenza: Creative Commons Attribution Non-commercial No Derivatives. Download (349kB) |
Abstract
Industry 5.0 is prompting the shift towards human-cyber-physical systems, where human operators and industrial collaborative robots (cobots) can work in synergy whilst using AI as a teammate instead of merely a tool. While Large Language Models (LLMs) offer intuitive, conversational interfaces for troubleshooting and cobot programming, their deployment in manufacturing environments is hindered by the risk of AI hallucinations and a lack of proprietary domain-related knowledge. The thesis presents an optimised Retrieval Augmented Generation (RAG) framework designed specifically to help operators in programming and troubleshooting industrial cobots. The specific case of the YASKAWA cobot in the Politecnico di Torino Mind4LAB is considered where its complex, unstructured and detailed manuals are utilised.
The research addresses the critical bottlenecks of data ingestion and retrieval using AI
Relatori
Anno Accademico
Tipo di pubblicazione
Numero di pagine
Corso di laurea
Classe di laurea
URI
![]() |
Modifica (riservato agli operatori) |
