Managing Functional Safety documentation represents a critical and resource-intensive activity for the development of systems compliant with ISO and IEC standards. This work documents the activities carried out during an internship at Bluewind s.r.l., aiming to evaluate the use of Large Language Models (LLMs) to optimize the creation and analysis of such documentation. During the internship, a generative AI-based application was developed, which will be integrated into the company workflow. The results demonstrate that, although LLMs can significantly accelerate documentation creation, human supervision remains essential to validate the technical correctness of the generated outputs and ensure compliance with regulatory safety requirements.
Gestire la documentazione per la Functional Safety rappresenta un'attività critica e onerosa per lo sviluppo di sistemi conformi agli standard delle normative ISO e IEC. Questo elaborato documenta l'attività svolta durante il tirocinio presso Bluewind s.r.l., il cui obiettivo era quello di valutare l'uso dei Large Language Models (LLM) per ottimizzare la creazione e l'analisi di tale documentazione. Durante il tirocinio è stato sviluppato un applicativo basato sull'intelligenza artificiale generativa che verrà integrato nel flusso di lavoro aziendale. I risultati ottenuti dimostrano che, sebbene gli LLM possano accelerare significativamente la creazione della documentazione, la supervisione umana rimane essenziale per validare la correttezza tecnica degli output generati e garantire la conformità ai requisiti di sicurezza normativi.
Applicazione dei Large Language Models ai processi di Sicurezza Funzionale
CORTESE, ALESSANDRO
2025/2026
Abstract
Managing Functional Safety documentation represents a critical and resource-intensive activity for the development of systems compliant with ISO and IEC standards. This work documents the activities carried out during an internship at Bluewind s.r.l., aiming to evaluate the use of Large Language Models (LLMs) to optimize the creation and analysis of such documentation. During the internship, a generative AI-based application was developed, which will be integrated into the company workflow. The results demonstrate that, although LLMs can significantly accelerate documentation creation, human supervision remains essential to validate the technical correctness of the generated outputs and ensure compliance with regulatory safety requirements.| File | Dimensione | Formato | |
|---|---|---|---|
|
Cortese_Alessandro.pdf
Accesso riservato
Dimensione
401.7 kB
Formato
Adobe PDF
|
401.7 kB | Adobe PDF |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/104206