National‑level "Pilot" Honor: AI‑Powered Steel Energy‑Consumption Optimization — Accurate Calculation and Substantial Energy Savings!
Recently, the results of the 2026 "Pilot Cup" Selection for High‑end Industrial Software Products and Typical Application Cases, hosted by the National Industrial Information Security Development Research Center, were officially announced. The innovative application of AI‑driven energy consumption management system for the steel industry, jointly submitted by Wuhan Iron and Steel Group Kunming Iron & Steel Co., Ltd. and Luculent Wisdom Technology Co., Ltd., has been successfully shortlisted in the Typical Cases for Intelligent Innovation and Development under the theme of AI + Industrial Software Industry Application Innovation. Only four achievements nationwide were selected for this category.
This selected case represents high‑level integration practices of domestic industrial software and artificial intelligence technologies with the real economy. Covering key industrial sectors including aerospace, steel, automobile and pharmaceuticals, the selection aims to identify replicable and promotable innovation paradigms, deliver demonstration‑driving effects, and boost high‑quality development of China’s industrial software industry.
In the past, dust removal operations at steel mills relied entirely on veteran operators. Wearing heavy masks, operators kept a close eye on instrument panels and manually adjusted fan frequencies. Excessively high fan output would waste massive electric energy, while insufficient output would easily cause excessive dust emissions. The whole operation depended entirely on personal experience. Today, dust removal processes are taken over by AI. Working conditions once governed solely by veteran operators’ experience are being redefined by algorithms.
Aligned with the "Dual‑Carbon" strategic goals, Luculent Sunshine AI Energy Consumption Management System addresses steel enterprises’ pain points of energy waste and extensive operation & maintenance. It connects the full‑link workflow of actual working conditions ‑ data collection ‑ AI decision‑making ‑ execution optimization. Acting as a seasoned energy‑saving dust‑removal expert, the system conducts real‑time second‑level perception, makes autonomous decisions at each working‑condition node, and keeps continuous optimization throughout operation cycles. It transforms high‑energy‑consumption equipment into energy‑saving profit‑driving assets for enterprises.
Energy consumption indicator prediction models and working‑condition optimization models are built to capture real‑time changes in production loads and automatically adjust equipment operating parameters. Digital twins are adopted for simulation verification of AI models. Once anomalies such as idling equipment or low energy efficiency are detected, the AI model triggers control actions. O&M personnel can also remotely adjust equipment via the IoT‑powered digital‑twin visualization interface to realize refined management.
The AI model is capable of reading operational trends, identifying optimal control parameters, and judging timing for parameter switching. Guided by precise algorithms, fans adjust speeds on demand, and full‑working‑condition energy‑saving diagnostics keep equipment operating along the golden efficiency curve, truly realizing AI‑driven regulation and intelligent dust removal.
Equipped with a dust video recognition model, the system functions like a pair of sharp eyes to monitor dust concentration round‑the‑clock with recognition accuracy above 95%. Combined with monitoring results, it dynamically adjusts air‑volume balance parameters. Acting as an industrial brain, it interprets full‑production‑workflow data in real time and dynamically generates optimal dust‑removal strategies. Dust‑related data is continuously fed back for AI model training, enabling near‑limit operation balancing environmental compliance and energy conservation. Timed automatic ash removal is also supported to clear ash buildup and blockages in pipelines.
Single‑screen overview with multi‑dimensional drill‑down capabilities. The energy‑efficiency optimization cockpit integrates full‑dimensional data including online energy‑efficiency monitoring, comprehensive energy‑saving analysis, and analysis of environmental‑protection‑indicator changes, delivering multi‑level, multi‑perspective dynamic dashboards.
Managers can retrieve key indicators with one click, including baseline vs. actual power consumption and unit‑consumption trend curves. Through intuitive chart comparison, they can trace energy‑consumption curves, equipment status and dust‑removal efficiency in real time. Data speaks for itself, energy efficiency becomes tangible, and production processes shift from black‑box operation to transparent and controllable management.
Electricity Savings: For a 2240 kW dust‑removal fan, the system cuts power consumption by 5%‑15% after deployment, saving nearly RMB 400 000 in electricity fees annually, without running extra equipment merely for safety guarantees.
Labor Savings: Traditional manual workflows requiring on‑site real‑time attendance and manual air‑volume adjustment are replaced by an intelligent mode with scheduled supervision and automatic regulation.
Time Savings: Adopting a lightweight transformation scheme, deployment can be completed in around two months.
Tangible and measurable benefits constitute the ultimate benchmark for evaluating platform value.
At present, this AI energy‑consumption optimization system has become a trusted digital‑intelligent partner for major steel groups including Baowu Group and Hangzhou Iron & Steel Group. It has been selected for the MIIT 2025 Intelligent Manufacturing "Challenge‑Oriented Project" and included in the China Iron and Steel Association’s Ultimate Energy Efficiency Capability List, with its industry benchmark effect continuously expanding.
Competition among factories of the future no longer hinges on equipment quantity. Instead, it centers on superior scheduling, less waste, higher efficiency and greener production capacity.
Luculent has over 20 years of experience in industrial digitalization, with rich first‑hand know‑how from industrial sites, in‑depth understanding of industrial equipment characteristics, production‑process rules and energy‑consumption link logic. Leveraging its self‑developed vertical large model for equipment and AI agents, the system conducts deep reinforcement learning and simulation deduction on real‑time production data to achieve self‑perception, self‑decision‑making, self‑optimization and self‑operation & maintenance. It delivers powerful AI‑driven momentum for the green and low‑carbon transformation of the steel industry during the 15th Five‑Year Plan period.