Company News

On October 23, the award ceremony of the 2026 “Data Elements ×” Competition National Finals, co-hosted by the National Data Administration and others, was held at the Qingdao International Conference Center. Gtrontec’s project “AI Agents: Autonomous Decision-Making Digital Employees for Manufacturing Scenarios Driven by Model-Data Integration” won the National Third Prize. The project previously won the Second Prize in the Hubei Division and became the only project from Hubei to enter the offline final evaluation of the national industrial manufacturing track. This award demonstrates Gtrontec’s comprehensive competitiveness in industrial data value conversion, industrial AI autonomous decision-making, and industrialization applications.

The “AI Agents: Autonomous Decision-Making Digital Employees for Manufacturing Scenarios Driven by Model-Data Integration” project focuses on core challenges in complex industrial sites, such as difficult correlation of multi-source data, difficult understanding of process logic, and difficult reliable execution of intelligent decisions. Relying on Gtrontec’s Octopus Brain industrial intelligent decision-making hub and Agent Space industrial-grade agent platform, the project integrates cross-system industrial data, process mechanisms, and expert experience; organizes business objects such as equipment, orders, and abnormal events and their relationships into an industrial ontology semantic model, providing a deterministic reasoning and execution basis for AI. Large models handle semantic understanding, task planning, and decision scheduling, while small models are trained based on scenario data and industrial mechanisms through machine learning and deep learning to form scenario-specific professional capabilities. Professional analysis methods and operating specifications are codified into callable Skills, collaborating with industrial software to complete on-site business execution, forming a complete capability system spanning active perception, intelligent decision-making, and collaborative execution.

In complex manufacturing sites, AI is always tested by real business results. A major breakthrough of this award-winning project is the deep integration of digital employees’ professional analysis capabilities with the deterministic execution mechanisms of industrial software, enabling digital employees to autonomously complete cross-system data extraction and verification, multi-dimensional correlation analysis, and abnormal root-cause localization, and to link with business systems for execution and handling via API/MCP, continuously track and verify results, forming a business closed loop from anomaly discovery to problem resolution. According to estimates, deploying digital employees on a single line delivers comprehensive economic benefits exceeding RMB 1 million, improves anomaly handling efficiency by over 75% on average, enables 24/7 uninterrupted work, and releases more than 25% of frontline staff’s transactional work hours. Industrial data thus evolves from a basis for auxiliary analysis into a key element driving production actions.

What truly demonstrates industrial-grade strength is strict assurance of execution reliability and security boundaries. Through large-small model collaboration and standardized workflows, the project balances complex task reasoning with consistency of execution results; at the same time, it introduces mechanisms such as manual confirmation at key nodes, tiered permission management, sandbox isolation for data analysis code, and security auditing, ensuring that digital employee operations are controllable and traceable.

Enabling digital employees to work steadily, be controllable, and be verifiable in industrial sites with high complexity and high reliability requirements is the core competitiveness of this project. This capability stems from Gtrontec’s continuously deepened AI Native strategy. Leveraging more than 30,000 models and engineering experience in processes, equipment, yield, and production scheduling accumulated from long-term deep cultivation in manufacturing, the company organizes different business objects such as equipment, orders, and abnormal events into an industrial ontology semantic model; Octopus Brain coordinates industrial ontology, large models, specialized models, digital employees, and industrial software to drive intelligent decision-making and on-site business execution. At the same time, through AI FDE, it goes deep into customer production lines, integrating business diagnosis, scenario validation, and effect feedback throughout the delivery process, and codifies validated models, skills, and business processes into reusable standardized capabilities. This forms an industrial AI two-way data flywheel in which scenarios continuously generate data, data drives model evolution, and intelligent capabilities feed back into business, continuously strengthening Gtrontec’s technical accumulation and scalable delivery advantages.

This award provides a practical sample from the core manufacturing site for unleashing the value of data elements. As industrial data deeply integrates with process knowledge, intelligent models, and business execution, the value of data elements will continue to expand. In the future, Gtrontec is using the engineering practice of industrial AI autonomous decision-making to explore an effective path for transforming data elements into advanced productivity, and to provide verifiable and replicable industrial experience for industrial AI to move toward autonomous factory governance.