Edge AI integration architecture for the internet of things [Uç yapay zekanın nesnelerin internetine entegrasyon mimarisi]

Alkanalka, Dilara and Seçen, Mert Sami and Yalçın, Neşenaz and Yıldız, Hüseyin Eren and Çimenoǧlu, Erdem and Helvacioǧlu, Turgut and Keküllüoğlu, Dilara (2026) Edge AI integration architecture for the internet of things [Uç yapay zekanın nesnelerin internetine entegrasyon mimarisi]. In: 34th Signal Processing and Communications Applications Conference (SIU), Istanbul, Turkiye

Full text not available from this repository. (Request a copy)

Abstract

The rapid development and current productive use of artificial intelligence models have introduced new integration scenarios and potential. In light of these developments, various opportunities have arisen for the use of generative models in industrial environments, including their potential contributions to automation. However, these opportunities raise questions about the usability and adaptability of generative models in industrial settings with limited device capabilities. This study proposes an intelligent system design for human-operator and artificial intelligence interaction. The necessary models for this system were evaluated using zero-shot and few-shot settings. As a main contribution, the industrial uses and interoperability of language models such as Gemma and Llama were evaluated. Our study contributes to the evaluation of dynamic assistants, such as generative artificial intelligence, as an industrial agent.
Item Type: Papers in Conference Proceedings
Uncontrolled Keywords: edge computing; natural language processing; small language models
Divisions: Faculty of Engineering and Natural Sciences
Depositing User: Dilara Keküllüoğlu
Date Deposited: 09 Sep 2026 11:52
Last Modified: 09 Sep 2026 11:52
URI: https://research.sabanciuniv.edu/id/eprint/54454

Actions (login required)

View Item
View Item