Predictive Maintenance · Real-time Analytics
Daxie Petrochemical needed a safer and more responsive way to inspect hazardous corridors carrying flammable and explosive materials because manual patrols left blind spots, delayed information updates, exposed workers to dangerous conditions, and made early risk detection difficult.
Reported impact
Daxie Petrochemical accelerated hazard response, increased patrol coverage and frequency, reduced worker exposure to dangerous environments, and established a reusable AI-based safety-inspection model for future plant expansion.
Under the broader CNOOC Refining program, Daxie Petrochemical and Baidu AI Cloud built an intelligent corridor-inspection system using Sugar BI, the Tiangong IoT Platform, digital-twin visualization, AI models, and a special explosion-proof robot. The front-end robot enters the corridor carrying visible-light, thermal-imaging, sound, gas, vibration, infrared, and laser sensors, then streams data back to a central safety-warning platform. That 'AI brain' analyzes people, equipment, and environmental conditions in real time, supports panoramic corridor views, safety-status monitoring, alert management, and emergency decision support, and continuously improves models through an edge-cloud closed loop. Digital-twin functions make inspection processes and results visible on a large screen, while the platform's soft-hard integrated design supports standardized long-term expansion.
Explosion-proof inspection robots collect visual, thermal, acoustic, gas, vibration, infrared, and laser signals in the corridor and return them to the central warning platform. AI models analyze the feeds, digital-twin views expose the corridor state to operators, and edge-cloud feedback is used to issue alerts, coordinate response, and improve models over time.
The architecture formed in 2021 was described as mature, advanced, soft-hard integrated, intelligently visual, and edge-cloud integrated (original: 2021年2月初…形成了成熟先进、软硬件结合、智能可视、边云一体的系统集成架构).
The case says hazard response was shortened to the minute level, response speed improved 5-8x, and inspection coverage expanded by 70% (original: 将隐患响应时间缩短至分钟级别，响应时间提升了5-8倍；巡检作业范围…扩大了70%).
The corridor project required integrated soft-hard intelligent capabilities, including IIC-grade explosion-proof hardware and software able to support digital-twin visualization and optimized detection models (original: 管廊智能巡检项目对完整的软硬件一体智能化产品能力要求极高，其中硬件需要达到IIC级防爆，而软件不仅要实现数字孪生可视化控制，还需要基于深度学习，运用和持续优化检测模型).
The front-end robot carries visible-light, thermal-imaging, and sound sensors and sends data back to the central brain for multi-modal AI analysis and warning (original: 搭载可见光、热成像、声音传感器等数据感知设备…回传至“中枢大脑”…通过多模AI分析技术实时监测与分析管廊内人员、设备以及环境的状态).
The system combines sound, gas, infrared, laser, visible-light, and vibration data to detect risks such as abnormal temperature and small gas leaks that are hard for human patrols to find (original: 可综合声音、气体、红外、激光、可见光、震动传感器的实时数据…尤其是不放过温度异常、气体微小泄露等人工难以发现的安全风险).
Review how this reference was scored and trace each normalized claim back to its documented source.
Scoring methodology
Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.
Traceability
Field coverage shows which facts each source supports. Dates are shown only when documented by the source.
Related
Baidu AI Cloud Sample May 14, 2025 China Longyuan Power Group Corporation Ltd. (龙源电力) Longyuan Power needed a scalable way to monitor far-flung wind farms and turbines, identify safety and equipment risks faster, and turn production data into a more centralized, intelligent operating model for… Inspection-model accuracy exceeds 95%. Equipment Status Recognition Algorithms · Personnel Safety Recognition Algorithms · Production Operations Intelligent Platform 
In-house Sample Est. · Jul 10, 2026 Aramco Aramco needs to turn extremely large and complex exploration, drilling, refining, logistics, and market datasets into faster, safer, and more reliable operational decisions while reducing uncertainty, downtime, and risk… five facilities recognized by the World Economic Forum as Lighthouse sites Dammam-7 · Ghawar-1 · Operations Co-Pilot 
Bain Sample Archived · May 26, 2026 TCL Photovoltaic Technology TCL Photovoltaic Technology's growth created more dispersed projects, more complex operating processes, and higher policy risk, so the company needed AI to improve efficiency, consistency, and control as it scaled. 10%-15% auditing process cost reduction AI service group · AI-powered recognition · Intelligent inspection 