A robot hand guide device includes a first connection flange configured to connect the robot hand guide device to a tool flange of a robot arm, a second connection flange configured to fasten a tool to be handled by the robot arm to the robot hand guide device, and a transfer apparatus configured to transfer forces and torques between the first connection flange and the second connection flange. A sensor device attached to the transfer apparatus is configured to detect forces and torques transferred via the transfer apparatus. The robot hand guide device further includes at least a first connection element connected to the first connection flange and configured for detachably connecting a guide handle of the robot hand guide device to the first connection flange.

Generating a robot control policy that regulates both motion control and interaction with an environment and/or includes a learned potential function and/or dissipative field. Some implementations relate to resampling temporally distributed data points to generate spatially distributed data points, and generating the control policy using the spatially distributed data points. Some implementations additionally or alternatively relate to automatically determining a potential gradient for data points, and generating the control policy using the automatically determined potential gradient. Some implementations additionally or alternatively relate to determining and assigning a prior weight to each of the data points of multiple groups, and generating the control policy using the weights. Some implementations additionally or alternatively relate to defining and using non-uniform smoothness parameters at each data point, defining and using d parameters for stiffness and/or damping at each data point, and/or obviating the need to utilize virtual data points in generating the control policy.

Generating a robot control policy that regulates both motion control and interaction with an environment and/or includes a learned potential function and/or dissipative field. Some implementations relate to resampling temporally distributed data points to generate spatially distributed data points, and generating the control policy using the spatially distributed data points. Some implementations additionally or alternatively relate to automatically determining a potential gradient for data points, and generating the control policy using the automatically determined potential gradient. Some implementations additionally or alternatively relate to determining and assigning a prior weight to each of the data points of multiple groups, and generating the control policy using the weights. Some implementations additionally or alternatively relate to defining and using non-uniform smoothness parameters at each data point, defining and using d parameters for stiffness and/or damping at each data point, and/or obviating the need to utilize virtual data points in generating the control policy.

A control system of a surgical robot arm, the surgical robot arm comprising a series of joints by which the configuration of that surgical robot arm can be altered and one or more force or torque sensors, each force or torque sensor configured to sense a force or torque at a joint of the series of joints, the control system being configured to control the configuration of the surgical robot arm to be altered in response to an externally applied force or torque by: receiving sensory data from the one or more force or torque sensors indicative of a sensed force or torque at a part of the surgical robot arm resulting from the externally applied force or torque; determining a position of the part of the surgical robot arm using a reference position, whereby the sensed force or torque would be compensated by moving the part of the surgical robot arm to the determined position; sending a command signal to the surgical robot arm to drive the part of the surgical robot arm to the determined position; and updating the reference position if the difference between the reference position and the determined position is greater than a threshold displacement.

A robot system causing a robot to operate in response to a handling force, wherein a position of the robot can be adjusted with higher accuracy. In one aspect of the present disclosure, a robot system includes a robot having a handle, a force sensor configured to detect a handling force applied to the handle, and an inching motion execution section configured to execute an inching motion of causing the robot to move by a movement amount determined in response to the handling force detected by the force sensor.

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for distributing skill templates for robotic demonstration learning. One of the methods includes receiving from the user device by a skill template distribution system, a selection of an available skill template. The skill template distribution system provides a skill template, wherein the skill template comprises information representing a state machine of one or more tasks, and wherein the skill template specifies which of the one or more tasks are demonstration subtasks requiring local demonstration data. The skill template distribution system trains a machine learning model for the demonstration subtask using a local demonstration data to generate learned parameter values.

A robot system includes a robot including leading end, base, and multi-articular arm, and circuitry that controls the atm to move the end based on motion control program specifying transition over time of target position and posture of the end, the transition including correction target portion starting and ending in the transition; controls the arm to move the end in response to guided manipulation applying external force to the robot while the circuitry controls the arm; obtains relative command information based on the target position and posture at start of the correction portion and specifying the target position and posture at points in the correction portion including start and end in the correction portion; and controls the arm to move the end from the position and posture based on the information, beginning at time when movement of the arm controlled by the circuitry in response to the manipulation has ended.

A robot system includes a robot including leading end, base, and multi-articular arm, and circuitry that controls the atm to move the end based on motion control program specifying transition over time of target position and posture of the end, the transition including correction target portion starting and ending in the transition; controls the arm to move the end in response to guided manipulation applying external force to the robot while the circuitry controls the arm; obtains relative command information based on the target position and posture at start of the correction portion and specifying the target position and posture at points in the correction portion including start and end in the correction portion; and controls the arm to move the end from the position and posture based on the information, beginning at time when movement of the arm controlled by the circuitry in response to the manipulation has ended.

The collaborative operation support device includes a display device including a display area; and a processor configured to detect, based on an image in which the operator or the robot is represented, a position of a section of the robot in the display area when the operator looks at the robot through the display area, the section associated with an operation mode of the robot specified by means of an input device; select, in accordance with the specified operation mode of the robot, display data corresponding to the specified mode among display data stored in a memory; and display the selected display data in the display area of the display device in such a way that the selected display data is displayed at a position that satisfies a certain positional relationship with the position of the section of the robot in the display area.

Learning to effectively imitate human teleoperators, even in unseen, dynamic environments is a promising path to greater autonomy, enabling robots to steadily acquire complex skills from supervision. Various motion generation techniques are described herein that are rooted in contraction theory and sum-of-squares programming for learning a dynamical systems control policy in the form of a polynomial vector field from a given set of demonstrations. Notably, this vector field is provably optimal for the problem of minimizing imitation loss while providing certain continuous-time guarantees on the induced imitation behavior. Techniques herein generalize to new initial and goal poses of the robot and can adapt in real time to dynamic obstacles during execution, with convergence to teleoperator behavior within a well-defined safety tube.