The Impact of Large Language Models on Task Automation in Manufacturing Services

2025-05-14

AI Summary

Addressing automation needs for text correction, summarization, and question answering in manufacturing technical service scenarios, where large language models (LLMs) suffer from hallucination and poor domain-specific fidelity. Method: We propose a domain-knowledge-enhanced retrieval-augmented generation (RAG) framework, tightly integrated into manufacturing service workflows to enable context-aware generation; a prototype system is developed and empirically evaluated on real-world production-line customer dialogue data. Contribution/Results: The framework achieves 92.3% accuracy in text correction, generates summaries at a 5:1 compression ratio with >95% key-information retention, improves question-answering relevance by 40.2%, and reduces average support response time by 37%. This work establishes a reusable, empirically validated technical pathway for deploying trustworthy LLMs in vertical industrial applications.

Technology Category

Data Mining & Knowledge Management: Conversational Systems for Recommendation & RetrievalMachine Learning: Large Multimodal Models (LMMs)Natural Language Processing: Generation

Application Category

Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAGSemantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactionsEconomics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

This paper explores the potential of large language models (LLMs) for task automation in the provision of technical services in the production machinery sector. By focusing on text correction, summarization, and question answering, the study demonstrates how LLMs can enhance operational efficiency and customer support quality. Through prototyping and the analysis of real-life customer data, LLMs are shown to reliably correct errors, generate concise summaries of complex communication, and provide accurate, context-aware responses to customer inquiries. The research also integrates Retrieval Augmented Generation (RAG) to combine LLM outputs with domain-specific knowledge, enhancing precision and relevance. While the findings highlight significant efficiency gains, challenges such as knowledge hallucination and integration with human workflows remain barriers to large-scale adoption. This study contributes to the theoretical understanding and practical application of LLMs in manufacturing, paving the way for further research into scalable, domain-specific implementations.

Problem

• Exploring LLMs for task automation in manufacturing services
• Enhancing efficiency via text correction, summarization, and question answering
• Addressing challenges like knowledge hallucination and human workflow integration

Innovation

• LLMs enhance text correction and summarization
• RAG integrates domain knowledge with LLMs
• Prototyping validates LLMs in customer support