A controller calculates a correction amount of a position of a robot 1 at a movement point in a first movement path, and drives the robot 1 in a second movement path obtained by correcting the first movement path. The controller includes a second camera configured to detect a shape of a part after a robot apparatus performs a task, and a variable calculating unit configured to calculate, based on an output of the second camera, a quality variable representing quality of a workpiece. When the quality variable deviates from a predetermined determination range, a determination unit of the controller determines that the position or an orientation of the robot 1 needs to be modified based on a correlation between the correction amount of the position in the first movement path and the quality variable.

This disclosure provides systems, methods, and apparatuses, including computer programs encoded on computer storage media, that provide for training, implementing, or updated machine learning logic, such as an artificial neural network, to model a manufacturing process performed in a manufacturing robot environment. For example, the machine learning logic may be trained and implemented to learn from or make adjustments based on one or more operational characteristics associated with the manufacturing robot environment. As another example, the machine learning logic, such as a trained neural network, may be implemented in a semi-autonomous or autonomous manufacturing robot environment to model a manufacturing process and to generate a manufacturing result. As another example, the machine learning logic, such as the trained neural network, may be updated based on data that is captured and associated with a manufacturing result. Other aspects and features are also claimed and described.

A controller calculates a correction amount of a position of a robot 1 at a movement point in a first movement path, and drives the robot 1 in a second movement path obtained by correcting the first movement path. The controller includes a second camera configured to detect a shape of a part after a robot apparatus performs a task, and a variable calculating unit configured to calculate, based on an output of the second camera, a quality variable representing quality of a workpiece. When the quality variable deviates from a predetermined determination range, a determination unit of the controller determines that the position or an orientation of the robot 1 needs to be modified based on a correlation between the correction amount of the position in the first movement path and the quality variable.

A method for joining concealed workpiece parts by an energy beam, wherein a lower workpiece part and an upper workpiece part are positioned relative to each other; the upper workpiece part contacts the lower workpiece part along a joining contour; the energy beam is directed onto an upper side of the upper workpiece part, moved along the joining contour by a controller, in order to join the upper workpiece part to the joining contour; an exploratory seam is produced on the upper work piece part, for detecting the joining contour; a detector detects a boundary at which a surface area of the upper work piece part borders a surface area of the upper work piece part which does have contact with the joining contour; the controller registers a position of the boundary and compares it with a target position of the boundary which is stored in the controller.

This disclosure provides systems, methods, and apparatuses, including computer programs encoded on computer storage media, that provide for training, implementing, or updated machine learning logic, such as an artificial neural network, to model a manufacturing process performed in a manufacturing robot environment. For example, the machine learning logic may be trained and implemented to learn from or make adjustments based on one or more operational characteristics associated with the manufacturing robot environment. As another example, the machine learning logic, such as a trained neural network, may be implemented in a semi-autonomous or autonomous manufacturing robot environment to model a manufacturing process and to generate a manufacturing result. As another example, the machine learning logic, such as the trained neural network, may be updated based on data that is captured and associated with a manufacturing result. Other aspects and features are also claimed and described.

This disclosure provides systems, methods, and apparatuses, including computer programs encoded on computer storage media, that provide for training, implementing, or updated machine learning logic, such as an artificial neural network, to model a manufacturing process performed in a manufacturing robot environment. For example, the machine learning logic may be trained and implemented to learn from or make adjustments based on one or more operational characteristics associated with the manufacturing robot environment. As another example, the machine learning logic, such as a trained neural network, may be implemented in a semi-autonomous or autonomous manufacturing robot environment to model a manufacturing process and to generate a manufacturing result. As another example, the machine learning logic, such as the trained neural network, may be updated based on data that is captured and associated with a manufacturing result. Other aspects and features are also claimed and described.