At the 2023 Schneider Electric Innovation Summit, Schneider Electric released EcoStruxureTM AI engine, an enterprise-level one-stop, scenario-based and open AI model production and operation platform, integrating expertise in energy management and automation into AI models. Provide low code and even zero code AI applications for users such as business experts and data analysts. The EcoStruxure AI engine provides structured design concepts that accelerate the implementation and iteration of AI models throughout their lifecycle, from data preparation to model production and operations, and integrates with existing systems and data to help users use AI technologies to better solve business problems, improve efficiency and innovate.

The EcoStruxure AI engine’s functions cover AI application store, data set management, AI modeling and model operation and maintenance, to achieve the full life cycle management of AI models, and support third-party models. It covers the five modeling processes required by enterprise customers to achieve the artificial intelligence industry, including data preparation, model training, model deployment, model reasoning and model monitoring in the AI model life cycle, and adopts cloud edge collaboration architecture, cloud training model, real-time reasoning by sending to the edge platform, and control all kinds of equipment to perform AI tasks.

IS220PPROS1B

IS220PPROS1B

IS220PPROS1B

The AI App Store provides applications such as predictive maintenance of equipment, optimization of equipment energy consumption, anomaly detection, predictive analysis of quality, virtual intelligent advisors, etc. Users can easily select various algorithms and models and apply them to their own data sets, thereby reducing the time and cost of developing new models.
AI models can be applied in intelligent manufacturing fields such as equipment management, quality improvement, and energy consumption optimization, such as real-time monitoring of welding quality data.
Data set management can view data characteristics, data trends, and data quality;
Automated machine learning enables users to quickly create AI models; Canvas modeling allows users to create AI models in drag-and-drop mode, code modeling supports Python modeling, and model operation provides model management and monitoring.

The EcoStruxure AI engine has a wide range of applicability and scalability, as well as high stability and reliability, while users can also customize and share their own algorithms and models according to their needs.

At present, the EcoStruxure AI engine has been applied in various segments of the scenario, Schneider Electric industry experts based on their own products to create a standard application suite, such as AI-based intelligent security, air compressor predictive maintenance, for different application scenarios has launched more than 20 industrial models, can be directly deployed out of the box. For specific scenario data, users can quickly pre-study based on the rich model templates built into the EcoStruxure AI engine to train customized field models with higher accuracy. In the building energy saving scenario, through the method of machine learning and operations research, Schneider Electric has accumulated more than 1,000 intelligent diagnosis solutions including air conditioning chillers, cooling pumps, fresh air units, boilers, hot water pumps and other equipment, to achieve more than 95% accuracy of cold capacity and energy consumption prediction, compared with traditional building automation control average energy saving increased by 15%.

At Schneider Electric Supply Chain China, AI-based solutions have become a powerful tool to improve production efficiency, control costs and protect the environment. The AI visual inspection project has successfully covered 10 factories in China, which can detect defects on the surface of products, greatly improve the quality of products, and effectively reduce the waste of resources and time due to product defects. AI laser predictive maintenance projects are based on machine learning algorithms, and through the analysis of large amounts of data, machine failures can be predicted in advance and preventive maintenance can be carried out, thereby improving production efficiency and reducing machine maintenance costs. Over the past year, these AI solutions have resulted in $4 million in direct cost savings for factories. Among them, Schneider Electric Yizhuang plant has reduced energy consumption by 10% in the past three years through energy-saving renovation and energy system optimization based on EcoStruxure AI engine.

With the continuous deepening of the application of AI technology in the real industry, more complex non-standard scenarios will emerge, requiring corresponding model support, and EcoStruxure AI engine is the optimal solution to this problem.