On December 10-11, the seventh Industrial Internet Competition National Finals, hosted by the National Industrial Information Security Development Research Center, was held in Beijing. Luculent Smart, with its self-developed "Large Model-based Intelligent Agent Solution for Coal Machinery Equipment Early Warning and Analysis," stood out from over 700 projects across four industry tracks: discrete manufacturing, process manufacturing, industrial intelligent applications, and emerging industries. Achieving the best score across all industries, it won first prize in the Leading Group of the National Finals!

This competition is a highly authoritative and influential top-tier event in China's industrial internet field. It has become a key vehicle for promoting industrial digital transformation, fostering industry-university-research-application collaboration, as well as a bellwether for industry technology and a gathering place for high-end talent.

This year's competition, themed "Connect Things, Aggregate Data, Empower Industries," focused on data element circulation and genuine industrial needs. It provided participating teams with access to high-value datasets from dozens of industries including steel, automotive, and energy. Its evaluation system covered the entire chain from data governance and model building to enabling effectiveness and market promotion results. Experts from industry authorities, scholars from research institutes, technical heads from leading enterprises, and capital institutions jointly assessed the entries, comprehensively examining the innovation and application value of the solutions.

Luculent's award-winning solution, centered on large AI model technology and driven by equipment operation and maintenance intelligent agents for remote diagnosis services of coal machinery, possesses multiple highlights and advantages:

Constructs an Equipment Fault Early Warning Model Library

Deeply integrates five types of multimodal diagnostic technologies including vibration, oil analysis, and current spectrum analysis. It covers 10 types of core equipment such as shearers, scraper conveyors, transfer machines, and crushers, achieving a model warning accuracy rate as high as 95%.

Co-creates High-Quality Datasets with Clients

Encompasses three types of datasets including small models, a coal machinery equipment operation and maintenance large model, and a knowledge base. This leverages the value of data assets, improves model warning accuracy, suppresses hallucinations in large models, and enhances the understanding of professional knowledge.

Forms a Multi-Agent Collaborative Mode

Incorporates 28 intelligent agents for tasks like intelligent data quality review, intelligent root cause localization of faults, and automatic diagnostic report generation. This significantly improves analysis efficiency, accuracy, and traceability, allowing large models to gradually replace manual work in the automatic analysis of complex coal machinery faults.

Builds a Cloud-Edge Collaborative Remote Equipment Operation and Maintenance Platform

Technologically, it achieves collaboration between data and models. Operationally, it achieves collaboration between maintenance strategies, operation services, and improvements in coal machinery product processes. It opens up the entire process closed-loop from device access and monitoring to operation and maintenance management, enabling a shift from "passive maintenance" to "active predictive maintenance," and from "human-led" to "intelligent agent-led + human supervision" modes.

This solution has already assisted group enterprises such as China National Coal Group, China Coal Technology & Engineering Group, Huadian Coal Industry, Shaanxi Coal Industry, and National Energy Group in transforming their coal machinery equipment operation and maintenance models. It has not only enhanced industry safety production capabilities, creating a safer, smarter, more comfortable, and greener working environment for practitioners but also significantly reduced operation and maintenance costs, generating economic and social benefits. It promotes the inheritance of knowledge and experience and product optimization, driving efficient collaborative development across the upstream and downstream industrial chain involving coal machinery manufacturers, mining companies, maintenance firms, and insurance companies.

In the future, the company will continue to strengthen the construction of the Luculent Suchang Industrial Internet Platform, deepen the application of the Suchang Yaoguang Industrial Large Model and intelligent agent technology in energy, manufacturing, and other fields. It aims to promote industry intelligent development through innovative practice and contribute its intellectual power to the improvement of the industrial internet ecosystem and talent cultivation.