EIM’S MESSAGE: Edge Learning: A New Paradigm for Industrial Automation
Stevan STANKOVSKI
Edge learning is emerging as a transformative paradigm in industrial automation, bridging the gap between traditional machine vision and deep learning. Unlike deep learning, which demands large datasets, high computational power, and specialized expertise, edge learning leverages pre-trained algorithms combined with rule-based vision tools to enable efficient, on-device data processing. This approach reduces complexity, minimizes training requirements, and empowers both beginners and experts to deploy AI-driven solutions directly on factory floors.
Nada Stajšić-Golijanin
This paper explores the potential application of artificial intelligence (AI) tools in automating parts of the claims processing workflow within the insurance industry in the Republic of Serbia. The motivation for this research stems from the fact that current, time-consuming processes in insurance rely exclusively on human support, carrying significant operational and reputational risks due to potential errors during processing. The paper focuses on the use of AI, specifically Optical Character Recognition (OCR) and Large Language Models (LLMs), for analyzing unstructured, handwritten documents and converting them into structured outputs suitable for further processing within existing information systems. As a case study, the results of processing a scanned, manually completed motor vehicle damage claim form were analyzed using widely available AI tools: ChatGPT, Copilot, and Gemini. The results indicate that Copilot (GPT-4) and Gemini successfully extracted handwritten text and structured it into a format suitable for further processing (JSON). Additionally, Gemini identified inconsistencies in the data, highlighting the potential of AI tools in detecting errors or indicators of fraudulent activity. The proposed approach highlights the transformative potential of AI in enhancing the efficiency, accuracy, and reliability of claims registration processes. These findings suggest that AI-driven automation could significantly reduce operational risks and improve service delivery within insurance companies.