EDUCATION > Projects

Istraživanje mogućnosti i ograničenja primjene velikih jezičnih modela u razvoju proizvoda


Data retrieved from the FSB repository on Dabar, on September 29 2025, https://repozitorij.fsb.unizg.hr/islandora/object/fsb%3A12209

Summary:

This thesis explores the possibilities and limitations of applying large language models (LLMs), such as ChatGPT, in product development. The aim of the research was to examine to what extent these tools can contribute to engineering processes, with a focus on tasks such as identifying user needs, patent searching, function formulation, concept generation and evaluation, as well as support during detailed design. The thesis begins with a literature review on the use of LLMs in product development. Subsequently, specific development tasks are identified and guidelines are provided on how to effectively use LLMs in each of them. The analysis is based on personal experiments and practical examples, highlighting the models’ limitations, such as inconsistencies in technical accuracy and a tendency to generalize. Special emphasis is placed on the importance of clear and detailed communication with the model, as well as an iterative approach. Following the development tasks, the application of LLMs is demonstrated through three product development examples. Finally, the thesis presents recommendations for engineers using these tools for the first time in the context of product development. The results show that, despite certain limitations, LLMs can significantly enhance efficiency and creativity in product development when used thoughtfully and with oversight.

Mentor:

Author:

Ivan Krpan

Year:

2025

Type:

Master thesis

PDF:

Ivan Krpan - Master thesis