AI in Electrical Equipment

Shenzhen Hello Tech Energy, maker of Jackery portable power stations, worked with AWS on a plug-in for its overseas customer service system that uses Claude 3 Sonnet on Amazon Bedrock to identify intent and draft replies from a Bedrock knowledge base. About five weeks after launch, agents sent generated replies directly for more than 30% of enquiries, at 1/20 of the cost of a third-party AI tool.

At its Chasseneuil plant in France, Schneider Electric adopted Cognex In-Sight cameras with the OneVision AI environment, widening inspection from 5 to 17 control areas and cutting cycle time from 400 ms to 200 ms. Cognex reports a 70x lower false bad rate and 30% faster integration of new inspection applications, and Schneider plans to roll the standard out across more factories.

Since 2023 LS ELECTRIC's Cheongju plant has used generative AI to turn a few real defect samples into thousands of synthetic defect images for training the computer vision on a robotic inspection arm for circuit breakers. False rejects of good products fell from 10% to 0%, and plant defects stand at 7 PPM.

Schneider Electric deployed an AI-powered control solution on h2e POWER's 20 kW solid oxide electrolyzer in Pune. It has run over 6,000 hours of stable operation, with the levelized cost of hydrogen down by up to 10% (around EUR 500,000 a year for a typical 10 MW plant) and predictive maintenance in place.

ABB embedded its generative AI solution into the ABB Ability Energy Management System, enabling operators to query energy usage, emissions drivers, and equipment performance through natural language, eliminating the need to manually compile reports across multiple dashboards.

CATL is building its fourth AI factory in Zhongzhou near Luoyang, with the first two construction phases already operational. According to the article, AI-powered Xiaomo humanoid robots perform end-of-line and DCR tests, and AI, 3D cameras and edge computing have made traditional inspection steps obsolete.

Schneider Electric says its EcoStruxure Automation Expert platform with Microsoft Azure AI has run h2e POWER's high-temperature solid oxide electrolysis for green hydrogen autonomously for more than 6,000 hours, cutting the levelized cost of hydrogen by up to 10%, which Schneider equates to about €500,000 a year for a typical 10 MW plant.

Schneider Electric has moved nearly 100 AI use cases into production. Its customer care centers have AI analyze and route about 7.5 million tickets a year, and a self-healing supply chain platform on adaptive machine learning and IoT reports a six-day cut in days-in-inventory, according to the company.