Low-voltage electrical apparatus are essential basic equipment for the development of new power systems.To address the constraints facing the industry in product sensing,complex protection and control,data interoperability,model training,and the engineering deployment and trustworthiness of large language models,literature synthesis,comparative case analysis,typical scenario analysis and life-cycle framework mapping are employed to review the foundations of digital transformation in the industry and to establish a full-life-cycle technological pathway encompassing intelligent design,intelligent manufacturing,intelligent products,and intelligent operation.The analysis indicates that industrial intelligence should be built on seamless data integration,supported by the fusion of mechanism-based models,data-driven models,and knowledge-based models.Large language models and industrial AI agents shall exert a cross-cutting effect across the whole life cycle,while standardization should provide guidance.These elements enable a closed-loop data system based on coordinated device-edge-cloud collaboration.Accordingly,a three-stage development pathway is proposed for the industry,ultimately leading to the adoption of digital twins,edge intelligence,domain-specific large language models,industrial AI agents,and software-defined products. During this transformation,the industry should advance the development of shared testbeds,security frameworks,and a standards system.
