# US 2021/0197369 A1 · Robot system and supplemental learning method

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2021/0197369 A1
- Document ID: US20210197369A1
- Status: Published application
- Kind: A1
- Application number: 17/058,770
- Filed: 2019-05-24
- Published: 2021-07-01
- Earliest claimed: 2018-05-25
- Assignees: Kawasaki Jukogyo Kabushiki Kaisha
- Applicants: Kawasaki Jukogyo Kabushiki Kaisha
- Inventors: Hitoshi HASUNUMA; Takuya SHITAKA; Takeshi YAMAMOTO; Kazuki KURASHIMA
- Web page: https://patentgrep.com/patent/US20210197369A1

## CPC

- B25J 9/163: {learning, adaptive, model based, rule based expert control}
- B25J 9/1633: {compliant, force, torque control, e.g. combined with position control}
- B25J 9/1656: {characterised by programming, planning systems for manipulators}
- B25J 9/1697: {Vision controlled systems}
- B25J 13/085: {Force or torque sensors (B25J13/082, B25J13/084 take precedence)}
- B25J 13/088: {with position, velocity or acceleration sensors}
- B25J 19/023: {including video camera means}
- B25J 9/1602: {characterised by the control system, structure, architecture}
- B25J 9/1679: {characterised by the tasks executed}
- B25J 9/1687: {Assembly, peg and hole, palletising, straight line, weaving pattern movement}
- B25J 13/08: by means of sensing devices, e.g. viewing or touching devices
- G05B 19/42: Recording and playback systems, i.e. in which the program is recorded from a cycle of operations, e.g. the cycle of operations being manually controlled, after which this record is played back on the same machine
- G05B 2219/36039: Learning task dynamics, process
- G05B 2219/36489: Position and force
- G05B 2219/40032: Peg and hole insertion, mating and joining, remote center compliance
- G05B 2219/40102: Tasks are classified in types of unit motions
- G05B 2219/40116: Learn by operator observation, symbiosis, show, watch
- G05B 2219/40153: Teleassistance, operator assists, controls autonomous robot
- G05B 2219/40223: If insertion force to high, alarm, stop for operator assistance
- G05B 2219/40391: Human to robot skill transfer
- G05B 2219/40496: Hierarchical, learning, recognition level controls adaptation, servo level
- B25J: MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES ({manipulators specially adapted for use in surgery A61B34/70; manipulators used in cleaning hollow articles B08B9/04}; manipulators associated with rolling mills B21B39/20; manipulators associated with forging machines B21J13/10; {manipulators associated with picking-up and placing mechanisms B23P19/007}; means for holding wheels or parts thereof B60B30/00; {vehicles with ground-engaging propulsion means, e.g. walking members B62D57/02, B62D57/032; devices for picking-up and depositing articles or materials between conveyors B65G47/90, B65G47/91; manipulators with gripping or holding means for transferring packages B65H67/065}; cranes B66C; {manipulators used in the protection or supervision of pipe-line installations F17D5/00; walking equipment adapted for nuclear steam-generators F22B37/006}; manipulators specially adapted for, or associated with, nuclear reactors G21C; {apparatus used for handling wafers during manufacture or treatment of semiconductor H10P72/50})
- G05B: CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS (systems for controlling or regulating non-electric variables G05D; systems for regulating electric or magnetic variables G05F; control devices or systems insofar as characterised by mechanical features only G05G)

## Abstract

A robot system includes a robot, state detection sensors to, a timekeeping unit, a learning control unit, a determination unit, an operation device, and an input unit, and an additional learning unit. The determination unit determines whether or not the work of the robot can be continued under the control of the learning control unit based on the state values detected by the state detection sensors to and outputs determination result. The additional learning unit performs additional learning of the determination result indicating that the work of the robot cannot be continued, the operator operation force, work state output by the operation device and the input unit, and timer signal output by the timekeeping unit.

## Claims

### Claim 1

A robot system comprising:
a robot working based on an operation command;
a state detection sensor detecting and outputting a state value that indicates progress of work of the robot;
a learning control unit outputting a calculation operation force based on the state value detected by the state detection sensor by using a model, the model being constructed by machine learning of a work state, a next work state associated with the work state, and at least one set of the state value and the operation force associated with the state value;
a determination unit outputting a determination result indicating whether or not the work of the robot can be continued under the control of the learning control unit based on the state value detected by the state detection sensor;
an operation device being operated by an operator, the operation device being detecting and outputting an operator operation force that is an operation force applied by the operator;
an input unit accepting an input of the work state by the operator and outputting the input;
a switching device converting either the operator operation force or the calculation operation force into the operation command and outputting the operation command based on the operator operation force and the calculation operation force and;
an additional learning unit additionally learning the work state, the next work state associated with the work state, and at least one set of the state value and the operation force associated with the state value and updating the model based on the determination result indicating that the work of the robot cannot be continued, the work state output by the input unit, the operator operation force output by the operation device, and the state value detected by the state detection sensor.

### Claim 2 (depends on 1)

The robot system according to claim 1,
wherein the additional learning unit calculates the next work state associated with the work state based on the state value and updates the model by additionally learning the work state, the next work state, the state value, and the operator operation force.

### Claim 3 (depends on 1)

The robot system according to claim 1,
wherein the input unit receives the input by the operator of the next work state associated with the input work state and outputs the input to the additional learning unit, and
wherein the additional learning unit additionally learns the work state, the next work state, the state value, and the operator operation force and updates the model.

### Claim 4 (depends on 1)

The robot system according to claim 1,
wherein when the work state is different from the next work state, the additional learning unit additionally learns the work state, the next work state, and a plurality of sets of the state value and the operator operation force associated with the state value and updates the model.

### Claim 5 (depends on 1)

The robot system according to claim 1,
wherein the switching device converts either the operator operation force or the calculation operation force into the operation command based on a setting signal for converting either the operator operation force or the calculation operation force and the switching device outputs the operation command.

### Claim 6 (depends on 1)

The robot system according to claim 1,
wherein the switching device includes a sensor that detects magnitude of the operator operation force output by the operation device, and
wherein the switching device converts either the operator operation force or the calculation operation force into the operation command based on the magnitude of the detected operator operation force and the switching device outputs the command operation.

### Claim 7 (depends on 1)

The robot system according to claim 1,
wherein the learning control unit interrupts the output of the calculation operation force based on the determination result indicating that the work of the robot cannot be continued, and
wherein the learning control unit resumes the output of the calculation operation force when the learning control unit determines that the additional learning is completed.

### Claim 8 (depends on 1)

The robot system according to claim 1, further comprising:
a notification unit outputting a notification signal based on the determination result indicating that the work of the robot cannot be continued; and
a display device that displays based on the notification signal.

### Claim 9 (depends on 8)

The robot system according to claim 8,
wherein the learning control unit calculates a similarity degree indicating a degree to which the current state value is similar to the specific work state in the model based on the state value detected by the state detection sensor and the learning control unit outputs the similarity degree, and
wherein the notification unit outputs the notification signal based on the similarity degree and the determination result indicating that the work of the robot cannot be continued.

### Claim 10 (depends on 1)

The robot system according to claim 1,
wherein the learning control unit calculates a similarity degree indicating a degree to which the current state value is similar to the specific work state in the model based on the state value detected by the state detection sensor and the learning control unit outputs the similarity degree, and
wherein the determination unit outputs the determination result based on the state value and the similarity degree.

### Claim 11 (depends on 8)

The robot system according to claim 8, further comprising
a certainty acquisition unit acquiring a certainty degree indicating a degree of certainty of estimation when the model estimates and outputs the calculation operation force according to the input data input to the model,
wherein the notification unit outputs the notification signal based on the certainty degree and the determination result indicating that the work of the robot cannot be continued.

### Claim 12 (depends on 1)

The robot system according to claim 1, further comprising
a certainty acquisition unit acquiring a certainty degree indicating a degree of certainty of estimation when the model estimates and outputs the calculation operation force according to the input data input to the model,
wherein the determination unit outputs the determination result based on the certainty degree.

### Claim 13 (depends on 1)

The robot system according to claim 1, further comprising
a progress acquisition unit acquiring a progress degree indicating that the work state of the robot realized by the calculation operation force output by the model corresponds to a degree of progress of the work of the robot,
wherein the determination unit outputs the determination result based on the progress degree.

### Claim 14 (depends on 1)

The robot system according to claim 1,
wherein when the additional learning unit determines that the work state input to the input unit is included in the model, the additional learning unit modifies an estimate standard of the work state in the model based on the state value detected by the state detection sensor.

### Claim 15 (depends on 1)

The robot system according to claim 1,
wherein when the additional learning unit determines that the work state input to the input unit is not included in the model, the additional learning unit registers the work state input to the input unit in the model based on the state value detected by the state detection sensor.

### Claim 16

(canceled)

### Claim 17

An additional learning method for a robot system, the robot system comprising:
a robot working based on an operation command;
a state detection sensor detecting and outputting a state value that indicates progress of work of the robot;
a learning control unit outputting a calculation operation force based on the state value detected by the state detection sensor by using a model, the model being constructed by machine learning of a work state, a next work state associated with the work state, and at least one set of the state value and the operation force associated with the state value;
an operation device being operated by an operator, the operation device being detecting and outputting an operator operation force that is an operation force applied by the operator; and
a switching device converting either the operator operation force or the calculation operation force into the operation command and outputting the operation command based on the operator operation force and the calculation operation force,
the method comprising:
a determination step outputting a determination result indicating whether or not the work of the robot can be continued under the control of the learning control unit based on the state value detected by the state detection sensor;
an input receiving step accepting input of the work state and the operator operation force from the operation device; and
an additional learning step additionally learning the work state, the next work state associated with the work state, and at least one set of the state value and the operation force associated with the state value and updating the model based on the determination result indicating that the work of the robot cannot be continued, the work state input in the input receiving step, the operator operation force input in the input receiving step, and the state value detected by the state detection sensor.

### Claim 18

A robot system comprising:
a robot working based on an operation command;
a state detection sensor detecting and outputting a state value that indicates progress of work of the robot;
a timekeeping unit outputting a timer signal at predetermined time intervals;
a learning control unit outputting a calculation operation force based on the state value detected by the state detection sensor and the timer signal by using a model, the model being constructed by machine learning of a work state, a next work state associated with the work state, and at least one set of the state value and the operation force associated with the state value;
a determination unit outputting a determination result indicating whether or not the work of the robot can be continued under the control of the learning control unit based on the state value detected by the state detection sensor;
an operation device being operated by an operator, the operation device being detecting and outputting an operator operation force that is an operation force applied by the operator;
an input unit accepting an input of the work state by the operator and outputting the input;
a switching device converting either the operator operation force or the calculation operation force into the operation command and outputting the operation command based on the operator operation force and the calculation operation force and;
an additional learning unit additionally learning the work state, the next work state associated with the work state, and at least one set of the state value and the operation force associated with the state value and updating the model based on the determination result indicating that the work of the robot cannot be continued, the work state output by the input unit, the operator operation force output by the operation device, the state value detected by the state detection sensor, and the timer signal.

### Claim 19 (depends on 18)

The robot system according to claim 18,
wherein the additional learning unit calculates the next work state associated with the work state based on the state value and updates the model by additionally learning the work state, the next work state, the state value, and the operator operation force.

### Claim 20 (depends on 18)

The robot system according to claim 18,
wherein the timekeeping unit outputs the timer signal based on a trigger signal and the timer signal is output at the predetermined time interval from the time when the trigger signal is received,
wherein the learning control unit outputs the trigger signal when starting the output of the calculation operation force, and
wherein the additional learning unit outputs the trigger signal when detecting the input of the operator operation force.
