A method, a non-transitory computer readable medium, and an apparatus for operating the robotic control system comprising a master apparatus (64) in communication with an input device (58, 60) having a handle (102) and a slave system (54, 74) having a tool (66, 67) having an end effector (73) whose position and orientation is determined in response to a current position and current orientation of the handle. The method involves producing a desired end effector position and orientation in response to a current position and orientation of the handle. The method involves causing the input device to provide haptic feedback that impedes translational movement of the handle, while permitting rotational movement of the handle and preventing movement of the end effector, when a rotational alignment difference between the handle and the end effector meets a disablement criterion. The method further involves re-enabling translational movement of the handle when the rotational alignment difference meets an enablement criterion.

A technique for training a neural network, including generating a plurality of input vectors based on a first plurality of task demonstrations associated with a first robot performing a first task in a simulated environment, wherein each input vector included in the plurality of input vectors specifies a sequence of poses of an end-effector of the first robot, and training the neural network to generate a plurality of output vectors based on the plurality of input vectors. Another technique for generating a task demonstration, including generating a simulated environment that includes a robot and at least one object, causing the robot to at least partially perform a task associated with the at least one object within the simulated environment based on a first output vector generated by a trained neural network, and recording demonstration data of the robot at least partially performing the task within the simulated environment.

A method of operating a robotic control system comprising a master apparatus in communication with an input device having a handle and a slave system having a tool having an end effector whose position and orientation is determined in response to a position and orientation of the handle. The method involves producing a desired end effector position and a desired end effector orientation of the end effector, in response to a current position and a current orientation of the handle. The method further involves causing the input device to provide haptic feedback that impedes translational movement of the handle, while permitting rotational movement of the handle and preventing movement of the end effector, when a rotational alignment difference between the handle and the end effector meets a first criterion. The method further involves re-enabling translational movement of the handle when the rotational alignment difference meets a second criterion.

Aspects of the generation of new robotic motion trajectories are described. In one embodiment, a new robot motion trajectory may be generated by gathering demonstrated motion trajectories, adapting the demonstrated motion trajectories into robot-reachable motion trajectories based on a joint space of a robot model, for example, and generating motion harmonics with reference to the motion trajectories. Further, one or more constraints may be specified for a new goal. The weights of the motion harmonics may then be searched to identify or generate a new motion trajectory for a robot, where the new motion minimizes discrepancy from the demonstrated motion trajectories and error due to the at least one constraint. In the new motion trajectory, the degree to which the constraints are satisfied may be tuned using a weight. According to the embodiments, new motion variants may be generated without the need to learn or review new demonstrated trajectories.

Utilization of user interface inputs, from remote client devices, in controlling robot(s) in an environment. Implementations relate to generating training instances based on object manipulation parameters, defined by instances of user interface input(s), and training machine learning model(s) to predict the object manipulation parameter(s). Those implementations can subsequently utilize the trained machine learning model(s) to reduce a quantity of instances that input(s) from remote client device(s) are solicited in performing a given set of robotic manipulations and/or to reduce the extent of input(s) from remote client device(s) in performing a given set of robotic operations. Implementations are additionally or alternatively related to mitigating idle time of robot(s) through the utilization of vision data that captures object(s), to be manipulated by a robot, prior to the object(s) being transported to a robot workspace within which the robot can reach and manipulate the object.

Utilization of user interface inputs, from remote client devices, in controlling robot(s) in an environment. Implementations relate to generating training instances based on object manipulation parameters, defined by instances of user interface input(s), and training machine learning model(s) to predict the object manipulation parameter(s). Those implementations can subsequently utilize the trained machine learning model(s) to reduce a quantity of instances that input(s) from remote client device(s) are solicited in performing a given set of robotic manipulations and/or to reduce the extent of input(s) from remote client device(s) in performing a given set of robotic operations. Implementations are additionally or alternatively related to mitigating idle time of robot(s) through the utilization of vision data that captures object(s), to be manipulated by a robot, prior to the object(s) being transported to a robot workspace within which the robot can reach and manipulate the object.

Utilization of user interface inputs, from remote client devices, in controlling robot(s) in an environment. Implementations relate to generating training instances based on object manipulation parameters, defined by instances of user interface input(s), and training machine learning model(s) to predict the object manipulation parameter(s). Those implementations can subsequently utilize the trained machine learning model(s) to reduce a quantity of instances that input(s) from remote client device(s) are solicited in performing a given set of robotic manipulations and/or to reduce the extent of input(s) from remote client device(s) in performing a given set of robotic operations. Implementations are additionally or alternatively related to mitigating idle time of robot(s) through the utilization of vision data that captures object(s), to be manipulated by a robot, prior to the object(s) being transported to a robot workspace within which the robot can reach and manipulate the object.

A method of operating a robotic control system comprising a master apparatus in communication with an input device having a handle and a slave system having a tool having an end effector whose position and orientation is determined in response to a position and orientation of the handle. The method involves producing a desired end effector position and a desired end effector orientation of the end effector, in response to a current position and a current orientation of the handle. The method further involves causing the input device to provide haptic feedback that impedes translational movement of the handle, while permitting rotational movement of the handle and preventing movement of the end effector, when a rotational alignment difference between the handle and the end effector meets a first criterion. The method further involves re-enabling translational movement of the handle when the rotational alignment difference meets a second criterion.

A method of operating a robotic control system comprising a master apparatus in communication with an input device having a handle and a slave system having a tool having an end effector whose position and orientation is determined in response to a position and orientation of the handle. The method involves producing a desired end effector position and a desired end effector orientation of the end effector, in response to a current position and a current orientation of the handle. The method further involves causing the input device to provide haptic feedback that impedes translational movement of the handle, while permitting rotational movement of the handle and preventing movement of the end effector, when a rotational alignment difference between the handle and the end effector meets a first criterion. The method further involves re-enabling translational movement of the handle when the rotational alignment difference meets a second criterion.

A robot movement teaching apparatus including a movement path extraction unit configured to process time-varying images of a first workpiece and fingers or arms of a human working on the first workpiece, and thereby extract a movement path of the fingers or arms of the human; a mapping generation unit configured to generate a transform function for transformation from the first workpiece to a second workpiece worked on by a robot, based on feature points of the first workpiece and feature points of the second workpiece; and a movement path generation unit configured to generate a movement path of the robot based on the movement path of the fingers or arms of the human extracted by the movement path extraction unit and based on the transform function generated by the mapping generation unit.