Adaptive predictor apparatus and methods
Apparatus and methods for training and operating of robotic devices. Robotic controller may comprise a predictor apparatus configured to generate motor control output. The predictor may be operable in accordance with a learning process based on a teaching signal comprising the control output. An adaptive controller block may provide control output that may be combined with the predicted control output. The predictor learning process may be configured to learn the combined control signal. Predictor training may comprise a plurality of trials. During initial trial, the control output may be capable of causing a robot to perform a task. During intermediate trials, individual contributions from the controller block and the predictor may be inadequate for the task. Upon learning, the control knowledge may be transferred to the predictor so as to enable task execution in absence of subsequent inputs from the controller. Control output and/or predictor output may comprise multi-channel signals.
1. A robotic system, comprising:
a memory comprising computer readable instructions stored thereon; and
at least one processor configured to execute the computer readable instructions to,
operate the robotic system in two modes comprising a learning mode and an operation mode, the operation mode being different from the learning phase, and
generate an output signal during the operation mode, the output signal comprises a combination of first and second signals, the first signal based at least in part on movement of the robotic system during the learning mode, and the second signal based at least in part on actions of the robotic system initiated by an operator at specific locations during the learning mode.
2. The robotic system of claim 1 , wherein the at least one processor is further configured to execute the computer instructions to,
switch between the operation mode and the learning mode based on receipt of an input signal from a user interface of the robotic system.
3. The robotic system of claim 1 , wherein the at least one processor is further configured to execute the computer readable instructions to,
associate, in the learning mode, movement of the robotic system towards a target location based on an association developed between an action tag and a sensory context, the action tag corresponding to an input signal received by the robot, the sensory context corresponding to data collected by at least one sensor coupled to the robot, and
wherein, the output signal is configured to cause the robotic system to perform a desired action in absence of a contemporaneous input signal, the output signal being based on the movement of the robotic system during the learning mode.
4. The robotic system of claim 3 , wherein,
the output signal is configured to cause the robotic system to perform a desired action in absence of a contemporaneous input signal, the output signal being based on the movement of the robotic system during the learning mode.
5. The robotic system of claim 4 , wherein,
the output signal generated during the operation mode comprises a selection between the first signal and a third signal, the first signal being based at least in part on movement of the robotic system during the learning phase, the third signal being based in part on the sensory context received and the contemporaneous input signal being received;
the at least one processor is further configured to execute the computer readable instructions to,
select the first signal if the contemporaneous input signal is absent, and
select the third signal upon receipt of the contemporaneous input signal.
6. The robotic system of claim 3 , wherein,
the action tag corresponds to motion of the robot, including left turns, right turns, forward motion, reverse motion, or a combination thereof; and
the sensory context comprises a user input configured to cause the robotic system to execute at least one of the motions, including left turn, right turn, forward motion, reverse motion, or a combination thereof.
7. The robotic system of claim 3 , wherein,
the action tag is based on actions of the robotic system effectuated by a user of the robotic system causing the robotic system to execute the movement.
8. A non-transitory computer readable storage medium comprising a plurality of computer readable instructions embodied thereon, which when executed by at least one processor configure the at least one processor to,
operate the robotic system in two modes comprising a learning mode and an operation mode, the operation mode being different from the learning phase; and
generate an output signal during the operation mode, the output signal comprises a combination of first and second signals, the first signal based at least in part on movement of the robotic system during the learning mode, and the second signal based at least in part on actions of the robotic system initiated by an operator at specific locations during the learning mode.
9. The non-transitory computer readable storage medium of claim 8 , wherein the at least one processor is further configured to execute the computer instructions to,
switch between the operation mode and the learning mode based on receipt of an input signal from a user interface of the robotic system.
10. The non-transitory computer readable storage medium of claim 8 , wherein,
associate, in the learning mode, movement of the robotic system towards a target location based on an association developed between an action tag and a sensory context, the action tag corresponding to an input signal received by the robotic system, the sensory context corresponding to data collected by at least one sensor coupled to the robotic system, and
wherein, the output signal is configured to cause the robotic system to perform a desired action in absence of a contemporaneous input signal, the output signal being based on the movement of the robotic system during the learning mode.
11. The non-transitory computer readable storage medium of claim 8 , wherein,
the output signal is configured to cause the robotic system to perform a desired action in absence of a contemporaneous input signal, the output signal being based on the movement of the robotic system during the learning mode.
12. The non-transitory computer readable storage medium of claim 11 , wherein,
the output signal generated during the operation mode comprises a selection between the first signal and a third signal, the first signal being based at least in part on movement of the robotic system during the learning phase, the third signal being based in part on the sensory context received and the contemporaneous input signal being received;
the at least one processor is further configured to execute the computer readable instructions to,
select the first signal if the contemporaneous input signal is absent, and
select the third signal upon receipt of the contemporaneous input signal.
13. The non-transitory computer readable storage medium of claim 10 , wherein,
the action tag corresponds to motions of the robotic system, including left turns, right turns, forward motion, reverse motion, or a combination thereof; and
the sensory context comprises a user input configured to cause the robotic system to execute at least one of the motions, including left turns, right turns, forward motion, reverse motion, or a combination thereof.
14. The non-transitory computer readable storage medium of claim 10 , wherein,
the action tag is based on actions of the robotic system effectuated by an operator of the robotic system causing the robotic system to execute the movement.
15. A method for maneuvering a robotic system, comprising:
operating the robotic system in two modes comprising a learning mode and an operation mode, the operation mode being different from the learning phase; and
generating an output signal during the operation mode, the output signal comprises a combination of first and second signals, the first signal based at least in part on movement of the robotic system during the learning mode, and the second signal based at least in part on actions of the robotic system initiated by an operator at specific locations during the learning mode.
16. The method of claim 15 , further comprising:
switching between the operation mode and the learning mode based on receipt of an input signal from a user interface of the robotic system.
17. The method of claim 15 , further comprising:
associating, in the learning mode, movement of the robotic system towards a target location based on an association developed between an action tag and a sensory context, the action tag corresponding to an input signal received by the robotic system, the sensory context corresponding to data collected by at least one sensor coupled to the robotic system, and
wherein, the output signal is configured to cause the robot to perform a desired action in absence of a contemporaneous input signal, the output signal being based on the movement of the robot during the learning mode.
18. The method of claim 15 , wherein the output signal is configured to cause the robotic system to perform a desired action in absence of a contemporaneous input signal, the output signal being based on the movement of the robot during the learning mode.
19. The method of claim 18 , wherein the output signal is generated during the operation mode comprises a selection between the first signal and a third signal, the first signal being based at least in part on movement of the robotic system during the learning phase, the third signal being based in part on the sensory context received and the contemporaneous input signal being received.
20. The method of claim 19 , further comprising:
selecting the first signal if the contemporaneous input signal is absent, and
selecting the third signal upon receipt of the contemporaneous input signal.
21. The method of claim 17 , wherein,
the action tag corresponds to motion of the robot, including left turns, right turns, forward motion, reverse motion, or a combination thereof; and
the sensory context comprises a user input configured to cause the robotic system to execute at least one of the motions, including left turn, right turn, forward motion, reverse motion, or a combination thereof.
22. The method of claim 17 , wherein the action tag is based on actions of the robotic system effectuated by a user of the robotic system causing the robotic system to execute the movement.
23. A robotic system, comprising:
a memory comprising computer readable instructions stored thereon; and
at least one processor configured to execute the computer readable instructions to,
operate the robotic system in two modes comprising a learning mode and an operation mode, the operation mode being different from the learning phase,
generate an output signal during the operation mode, the output signal comprises a combination of first and second signals, the first signal based at least in part on movement of the robotic system during the learning mode, and the second signal based at least in part on actions of the robotic system initiated by an operator at specific locations during the learning mode,
switch between the operation mode and the learning mode based on receipt of an input signal from a user interface of the robotic system,
associate, in the learning mode, movement of the robotic system towards a target location based on an association developed between an action tag and a sensory context, the action tag corresponding to an input signal received by the robot, the sensory context corresponding to data collected by at least one sensor coupled to the robot, and
wherein, the output signal is configured to cause the robotic system to perform a desired action in absence of a contemporaneous input signal, the output signal being based on the movement of the robotic system during the learning mode.