Industrial AI has entered a critical window for large-scale deployment. Penetrating every link covering equipment, production lines, energy management, safety supervision, operation and maintenance with unprecedented breadth and depth, it is reshaping the operational paradigms, value creation mechanisms and business models of the industrial sector.

Many enterprises fall into the trap of "prioritising technological R&D over practical implementation" when rolling out industrial AI solutions, failing to deliver tangible business value. With 27 years of deep engagement in the industrial sector, Luculent has developed a comprehensive implementation methodology. It has built a fully-stacked, self-evolving industrial AI system featuring five interlinked layers: Talent Pool – Technology Platform – Digital Intelligence – Business Enablement – Collaborative Innovation. This system fuels the transformation of process industries and advances high-quality development of the new-type industrialisation drive.

I. Talent Foundation: Two Institutes & One Research Centre Address the Core Pain Point of "AI Lacking Industrial Domain Expertise"

To enable AI to deliver practical value on the factory floor, a workforce well-versed in manufacturing operations is indispensable.

Luculent has established two research institutes and one central R&D centre: the Smart Energy Research Institute, the Industrial Internet Research Institute, and the Platform R&D Centre. It has assembled a cross-functional AI team comprising IT, operational technology (OT), algorithm and model development specialists. With decades of professional experience in power generation, coal mining, iron and steel, cement and other sectors, these industry experts boast profound domain insights and extensive engineering practice.

Hundreds of specialists are stationed on-site at project locations across the country to decode production processes, analyse equipment operating conditions, and accumulate massive industrial know-how, which is further encapsulated into industrial models. Endowed with native industrial expertise, these solutions are highly tailored to production line scenarios. The team continuously delivers field-proven, quick-return practical experience to drive iterative upgrades of AI technologies.

II. Platform Backbone: National Cross-Industry Cross-Region Platform Delivers One-Stop AI Services

As the core infrastructure underpinning all AI capabilities, the Suchang Industrial Internet Platform has been listed in the Ministry of Industry and Information Technology’s (MIIT) Cross-Industry and Cross-Region Industrial Internet Platform roster for four consecutive years. In 2025, it was rated Grade A (the highest tier) and ranked among the nation’s top 10 platforms.

The company keeps ramping up R&D investment in five core technologies: industrial real-time data collection and processing, component-based platform development, and digital twins. It connects five hierarchical layers: equipment, production line, individual factory, enterprise and group headquarters, delivering one-stop "AI Platform + Scenario Application" services that cover the full toolkit for model training, intelligent agent orchestration and application development.

AI technologies are encapsulated as plug-and-play cloud-native services, eliminating the need for enterprises to build technical architectures and dedicated teams from scratch. Businesses can leverage standardised, professional AI tools to embark on intelligent upgrading with lower barriers and higher efficiency.

III. Digital Intelligence Core: Five Core Capabilities Close the Loop from Data Analysis to Automated Control

Data serves as the fuel for AI, industrial models act as decision-making brains, and automated regulation holds the key to unlocking commercial value.

1. High-Quality Dataset DevelopmentWe streamline data processing and modelling workflows and build extensive industry knowledge corpora. Seven high-quality datasets have been jointly developed with clients, among which the coal machinery equipment dataset has been included in MIIT’s pilot catalogue for high-quality industrial datasets, optimising the full lifecycle of equipment R&D, manufacturing and maintenance.

2. Industrial Small Model DevelopmentThe Tianji Data Intelligence Platform embeds a vast library of industrial algorithm functions, mature scenario-based modelling workflows and model optimisation tools. Equipped with over 100 industrial computational functions and 20 frequently deployed scenario models (including intelligent equipment early warning and energy consumption optimisation), it accelerates the rollout of intelligent applications across diverse industries.

3. Vertical Domain Large Model DevelopmentThe Suchang Yaoguang Industrial Large Model focuses on the power and coal mining sectors, featuring 155 industrial intelligent agents that cover ten core business domains including equipment maintenance and workplace safety. Its technological achievements have passed professional evaluation by the China Academy of Information and Communications Technology (CAICT) and won the first prize in the industrial track of the 7th National Industrial Internet Innovation Competition.

4. Closed-Loop Optimisation & ControlWe have pioneered a collaborative engine integrating large-model decision-making and small-model parameter optimisation. Analytical insights generated by AI can be directly deployed to DCS and PLC systems to realise automatic regulation, forming a closed workflow of fault early warning, intelligent diagnosis, scheme optimisation and automated execution. Taking energy efficiency optimisation for steel dust removal fans as an example, a single piece of equipment can save 400,000 kWh of electricity annually, marking a shift from passive data monitoring to proactive energy-saving automatic optimisation.

5. Global Cloud-Edge-End SynergyA cloud-edge-end integrated industrial brain enables three-level collaborative sharing of data, models and applications. The benchmark State Power Investment thermal power management platform achieves end-to-end digital connectivity across the group headquarters, 12 power plants and 37 generating units. It delivers a fault early warning accuracy rate exceeding 90%, boosts equipment operational reliability by 10%, and cuts coal consumption by 0.45 grams per kWh of power generated.

IV. Business Application Reach: Full-Lifecycle Digital Services Empower Over a Thousand Manufacturing Facilities

The ultimate goal of AI capabilities is to solve on-site production challenges and generate tangible economic returns.

Luculent provides enterprises with digital transformation consulting and integrated investment, construction, operation and maintenance services, alongside industry-specific smart solutions for power plants, coal mines, metallurgy, building materials and other sectors. To date, we have served more than 1,000 large central state-owned enterprises including China’s five major power generation groups, Baowu Steel, China National Coal Group and Anhui Conch Cement. Centred on core objectives of cost reduction, quality improvement, efficiency enhancement and safety assurance, we have developed replicable digital transformation benchmarks that translate AI strengths into measurable economic gains.

V. Innovation Ecosystem Fertile Ground: Building an Open Industrial Ecosystem Aligned with National Strategies

Sustainable iteration of industrial AI systems relies on open innovation and collaborative industrial layout.Luculent aligns its development with national strategies including AI-enabled advanced manufacturing and dual-carbon goals, venturing into emerging sectors such as integrated energy management, carbon asset administration and electricity trading.

We pioneer a collaborative model between central state-owned enterprises and private firms: central SOEs contribute industrial scenarios, data and computing resources, while private enterprises bring agility and innovative capacity to jointly advance technological breakthroughs and innovative business models. We export our end-to-end AI platforms and technical capabilities to digital technology subsidiaries of central SOEs.

We partner with leading industrial players and industry associations to formulate unified industry standards and foster a robust industrial ecosystem. Our globalisation strategy has enabled the export of self-developed domestic industrial software and industrial AI solutions to more than 20 countries across four continents.

Rooted in frontline industrial expertise, underpinned by a national-level industrial internet platform, driven by closed-loop digital intelligence capabilities, scaled via widespread industry deployment and sustained through open collaborative innovation, Luculent’s replicable, self-evolving full-stack industrial AI system accelerates the large-scale adoption of industrial AI and facilitates the high-end, green transformation of traditional industrial value chains.