利索-全球专利检索

  1. 登录
  2. 注册

搜索结果

11 条记录 (耗时 0.011 秒)

    SIMULATION ASSISTED PLANNING OF MOTIONS TO LIFT HEAVY OBJECTS
    2.
    发明公开
    SIMULATION ASSISTED PLANNING OF MOTIONS TO LIFT HEAVY OBJECTS 审中-公开

    公开(公告)号：EP3613545A1

    公开(公告)日：2020-02-26

    申请号：EP18190702.3

    申请日：2018-08-24

    申请人： Siemens Aktiengesellschaft

    发明人： MCDANIEL, Richard Gary

    IPC分类号： B25J9/16

    摘要： According to other embodiments, a method planning of motions to lift heavy objects using a robot system comprising a robot and an end effector, includes identifying data comprising (a) rigid bodies included in the robot and the end effector, (b) joints connecting the rigid bodies, and (c) torque limits for each of the joints. The torque limit for a joint indicates a maximum supported torque by a drive operating the joint. A motion path searching algorithm is applied to the input data to identify feasible robot paths. The motion path searching algorithm determines torque of each of joint when evaluating points for inclusion in a feasible robot path. An evaluated point is only included in a feasible robot path if the torque of each of the joints do not exceed the torque limits. At least one of the feasible robot paths is transferred to a controller associated with the robot.

    GENERATING LEARNED KNOWLEDGE FROM AN EXECUTABLE DOMAIN MODEL
    3.
    发明公开
    GENERATING LEARNED KNOWLEDGE FROM AN EXECUTABLE DOMAIN MODEL 审中-公开

    公开(公告)号：EP3542971A3

    公开(公告)日：2019-12-25

    申请号：EP18214015.2

    申请日：2018-12-19

    申请人： Siemens Aktiengesellschaft

    发明人： MCDANIEL, Richard Gary ,  QUIROS ARAYA, Gustavo Arturo ,  FRADKIN, Dmitriy

    IPC分类号： B25J9/16

    摘要： A computer-implemented method for performing autonomous operations in an operating environment includes simulating the operating environment to generate a plurality of examples. Each example comprises (a) signal data describing a scene, (b) one or more objects present in the scene, and (c) a description of characteristics associated with the objects. A machine learning model is trained using the examples to generate a data structure comprising (i) objects associated with a signal and (ii) characteristics corresponding to the objects associated with the signal. A signal sensor of an autonomous device collects an input signal describing a new scene. The machine learning model is used to generate an output data structure based on the input signal. One or more objects are identified using the output data structure. One or more actions are generated for operating the autonomous device based on the characteristics associated with the identified objects.