banner

SPIC Guizhou Jinyuan and Luculent Co-develop a Safety Agent

Industry trends 2026.11.02

On October 27, the 2026 Guizhou Provincial AI Industry LLM Achievements Release and Supply-Demand Matching Event was held in Guiyang. Witnessed by leading officials from the Guizhou Big Data Development Administration and relevant competent authorities, Luculent, as a vertical LLM technology provider, signed a cooperation agreement for scenario development and application of the Agent for Semantic Parsing of Safety Regulations and High-Risk Work Plans with Intelligent Generation of Safety Measures with SPIC Guizhou Jinyuan Group.

This Agent will equip Guizhou Jinyuan’s AI-enabled Penetrative Supervision Platform for Safety, Environmental Protection and Occupational Health with advanced AI capabilities. Leveraging LLM semantic understanding and video intelligent perception technologies, it establishes an intelligent review and on-site dynamic verification system for “two permits plus four measures and two plans”. The system covers intelligent review of work permits and special plans, on-site execution comparison and automatic early warning of risk deviations, enabling full-process intelligent management and control.

In terms of management and control framework, it supports pre-event intelligent compliance review, in-event dynamic video verification and post-event deviation tracing and review. It can accurately identify inconsistencies between permit plans and on-site implementation as well as inadequate implementation of safety measures, forming a closed loop of “intelligent review of permit plans — on-site dynamic verification — deviation early warning”. This comprehensively improves the standardization, intelligence and refinement of high-risk operation management.

For compliance, it is fully aligned with national laws and regulations, industrial codes and corporate safety rules. Every safety measure intelligently generated can be traced back to corresponding provisions. It addresses pain points of the conventional manual preparation mode, such as heavy reliance on experience, vague descriptions, missing key points and mismatched scenarios. It is expected to achieve a compliance rate of safety measures exceeding 98% and a 90% reduction in missing items for high-risk work measures, advancing standardized and accurate delivery of safety measures.

In terms of efficiency, the preparation cycle of initial plans is expected to be cut by over 70%, and the workload of safety regulation comparison reduced by 80%. Safety managers can be freed from repetitive work including regulation retrieval and drafting of measures item by item, so that they can focus on high-value work such as major risk assessment and dynamic on-site supervision.

Going forward, Luculent will continue to deepen the integration of AI and the energy sector, strengthen the safety production defense line with industrial intelligence, and inject strong momentum into the high-quality development of the digital economy in Guizhou Province.


Return to List Page
Share on WeChat Moments ×
Open WeChat and click on "Discover" at the bottom,
Use 'Scan' to share web pages on your social media circle。