Generating a robot control policy from demonstrations collected via kinesthetic teaching of a robot
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.
1. A method implemented by one or more processors, comprising:
receiving a group of data points generated based on sensor data from one or more sensors of a robot during physical manipulation of the robot, the physical manipulation being by a user to traverse a reference point of the robot from an initial point to a target point;
for each of the data points of the group:
assigning only d stiffness parameters to the data point, wherein d is a dimension of a task associated with the physical manipulation of the robot,
wherein the stiffness parameters assigned to each of the data points are based on sensor data, from a pressure sensor of the robot, during the physical manipulation,
wherein the sensor data is based on force applied to the pressure sensor during the physical manipulation of the robot, and
wherein the stiffness parameters assigned to a given data point, of the data points, are inversely proportional to spatial variance of the data points near the given data point;
generating a control policy that regulates both robot motion and robot interaction with an environment, wherein generating the control policy comprises using the data points and the stiffness parameters for the data points in learning a potential function for use in the control policy, the potential function having a global minimum based on the target point; and
controlling the robot, or an additional robot, based on the control policy.
2. The method of claim 1 , further comprising identifying the task based on user input through a user interface input device.
3. The method of claim 1 , further comprising identifying the task based on user input through a user interface input device.
4. The method of claim 1 , further comprising identifying the task based on the physical manipulation of the robot.
5. The method of claim 1 , further comprising identifying the task based on one or more objects detected by camera sensors during the physical manipulation of the robot.
6. A system, comprising:
a robot;
one or more computers comprising:
memory storing instructions;
one or more processors operable to execute the instructions to:
receive a group of data points generated based on sensor data from one or more sensors of the robot during physical manipulation of the robot, the physical manipulation being by a user to traverse a reference point of the robot from an initial point to a target point;
for each of the data points of the group:
assign only d stiffness parameters to the data point, wherein d is a dimension of a task associated with the physical manipulation of the robot,
wherein the stiffness parameters assigned to each of the data points are based on sensor data, from a pressure sensor of the robot, during the physical manipulation,
wherein the sensor data is based on force applied to the pressure sensor during the physical manipulation of the robot, and
wherein the stiffness parameters assigned to a given data point, of the data points, are inversely proportional to spatial variance of the data points near the given data point;
generate a control policy that regulates both robot motion and robot interaction with an environment, wherein generating the control policy comprises using the data points and the stiffness parameters for the data points in learning a potential function for use in the control policy, the potential function having a global minimum based on the target point; and
control the robot based on the control policy.
7. The system of claim 6 , wherein one or more of the processors are further operable to execute the instructions to identify the task based on user input through a user interface input device.
8. The system of claim 6 , wherein one or more of the processors are further operable to execute the instructions to identify the task based on user input through a user interface input device.
9. The system of claim 6 , wherein one or more of the processors are further operable to execute the instructions to identify the task based on the physical manipulation of the robot.
10. The system of claim 6 , wherein one or more of the processors are further operable to execute the instructions to identify the task based on one or more objects detected by camera sensors during the physical manipulation of the robot.