REAL-TIME PRODUCTION THROUGHPUT PREDICTION WITH MACHINE LEARNING

    公开(公告)号:US20240144108A1

    公开(公告)日:2024-05-02

    申请号:US17979645

    申请日:2022-11-02

    申请人: Hitachi, Ltd.

    发明人: Lili ZHENG

    IPC分类号: G06N20/20

    CPC分类号: G06N20/20

    摘要: In example implementations described herein, systems and methods are presented for predicting a throughput for a production line which can involve collecting a current set of production line data associated with a production line; and computing a predicted production throughput for the production line over a production period based on a boosted gradient decision tree algorithm applied to the current set of production line data, wherein the boosted gradient decision tree algorithm is trained on historical production line data.

    PRODUCTION LINE ANALYZER AND VISUALIZER FOR LINE PERFORMANCE IMPROVEMENT

    公开(公告)号:US20240142344A1

    公开(公告)日:2024-05-02

    申请号:US17976630

    申请日:2022-10-28

    申请人: Hitachi, Ltd.

    IPC分类号: G01M99/00

    CPC分类号: G01M99/005

    摘要: In example implementations described herein, there are systems and methods enabling manufacturers to become more efficient by providing a digital tool to track, visualize, quantify, and notify production bottlenecks for quick actions. Some implementations include an apparatus including a processor configured to analyze time-stamp data collected associated with a plurality of elements of an industrial process. The processor may be configured to generate, based on the analyzed time-stamp data, a color coded visualization of the industrial process, where generating the color coded visualization includes associating each of the plurality of elements of the industrial process with a corresponding color indicating a state of an element in the plurality of elements of the industrial process. The processor may further be configured to present, via a display, the generated color coded visualization of the industrial process.