Robot, information processing system, and storage medium

Patent metadata

Field	Value
Publication number	US-10059004-B2
Application number	US-201615176799-A
Country	US
Kind code	B2
Filing date	Jun 8, 2016
Priority date	Jun 22, 2015
Publication date	Aug 28, 2018
Grant date	Aug 28, 2018

Abstract

A robot includes a three-dimensional shape detecting sensor to detect a three dimensional shape of a travel surface existing in a forward travelling direction of the robot, a posture stabilizer to stabilize a posture of a body of the robot, a feature data generator to generate feature data of the detected three dimensional shape, an inclination angle prediction generator to generate a prediction value of an inclination angle of the body when the robot is to reach a position on the travel surface in the forward travelling direction at a future time point based on the feature data and a prediction model, and an overturn prevention controller to control the posture stabilizer to prevent an overturn of the robot based on the prediction value.

First claim

What is claimed is:

1. A robot comprising:
circuitry configured to:
detect a three dimensional shape of a travel surface existing in a forward travelling direction of the robot;
stabilize a posture of a body of the robot;
generate feature data of the detected three dimensional shape;
generate a prediction value of an inclination angle of the body when the robot is to reach a position on the travel surface in the forward travelling direction at a future time point based on the feature data and a prediction model; and
control to prevent an overturn of the robot based on the prediction value, wherein the circuitry controls a travelling speed of the robot so that a first time when the robot reaches the position on the travel surface in the forward travelling direction of the robot becomes longer than a second time required to complete stabilizing the posture of the body of the robot without causing the overturn of the robot.

2. The robot of claim 1 , wherein the prediction model employs a regression model that uses the feature data of the three dimensional shape of the travel surface as an explanatory variable, and the inclination angle of the body when the robot is to reach the position on the travel surface as a response variable.

3. The robot of claim 1 , wherein the circuitry is further configured to:
detect the inclination angle of the body; and
correct the prediction value based on the inclination angle detected.

4. The robot of claim 3 , wherein the circuitry is further configured to control driving of the body based on the prediction value or the corrected prediction value to change the travelling direction of the robot.

5. The robot of claim 3 , wherein the circuitry is further configured to:
generate a combination of the feature data of the three dimensional shape of the travel surface generated and the inclination angle detected when the robot is to reach the position on the travel surface in the forward travelling direction at the future time point as learning data; and
generate the prediction model by applying machine learning using the generated learning data.

6. The robot of claim 1 , wherein the inclination angle includes a roll angle and a pitch angle.

7. The robot of claim 1 , wherein the circuitry is further configured to:
store the prediction value using a first-in first-out (FIFO) queue with a number “n” settable with a whole number of two or more,
wherein the circuitry detects the three dimensional shape of the travel surface ahead of the robot at a distance of “(n+1)×Δd” each time the robot travels for a given distance “Δd,”
wherein the circuitry generates the prediction value of the inclination angle of the body each time the robot travels for the given distance “Δd,” and stores the prediction value in the FIFO queue,
wherein the circuitry reads out the prediction value from the FIFO queue each time the robot travels for the given distance “Δd,” and controls the stabilization of the posture of the body of the robot based on the read-out prediction value.

8. A non-transitory storage medium storing a program that, when executed by a computer, causes the computer to execute a method of stabilizing a posture of a body of a robot, the method comprising:
detecting a three dimensional shape of a travel surface in a forward travelling direction of the robot by using a three-dimensional shape detecting sensor;
generating feature data of the detected three dimensional shape;
generating a prediction value of an inclination angle of the body when the robot is to reach a position on the travel surface in the forward travelling direction based on the generated feature data and a prediction model, the prediction model using the feature data of the three dimensional shape of the travel surface as an explanatory variable and the inclination angle of the body when the robot is to reach the position on the travel surface in the forward travelling direction at a future time point as a response variable; and
controlling a posture stabilizer based on the prediction value to prevent an overturn of the robot by controlling a travelling speed of the robot so that a first time when the robot reaches the position on the travel surface in the forward travelling direction of the robot becomes longer than a second time required to complete stabilizing the posture of the body of the robot without causing the overturn of the robot.

9. An information processing system comprising:
a robot being one or more robots; and
an information processing apparatus connectable with the robot via a network,
the robot including first circuitry configured to:
detect a three dimensional shape of a travel surface existing in a forward travelling direction of the robot;
stabilize a posture of a body of the robot;
generate feature data of the detected three dimensional shape, and transmit the generated feature data to the information processing apparatus;
control to prevent an overturn of the robot;
detect an inclination angle of the body; and
generate a combination of the feature data of the three dimensional shape of the travel surface generated and the inclination angle detected when the robot is to reach a position on the travel surface in the forward travelling direction at a future time point as learning data, and transmits the generated learning data to the information processing apparatus,
the information processing apparatus including second circuitry configured to:
generate a prediction value of the inclination angle of the body when the robot is to reach the position on the travel surface in the forward travelling direction at the future time point based on the feature data received from the robot and a prediction model, and transmit the generated prediction value to the robot; and
generate the prediction model by applying machine learning using the generated learning data,
wherein the first circuitry of the robot controls to prevent an overturn of the robot based on the prediction value received from the information processing apparatus.

Assignees

• Inaba Daiki
• Hatanaka Wataru
• Shimura Hiroshi
• Kawaguchi Atsuo
• Ricoh Co Ltd

Inventors

• Inaba Daiki
• Hatanaka Wataru
• Shimura Hiroshi
• Kawaguchi Atsuo

Classifications

• B25J5/00

  Manipulators mounted on wheels or on carriages (B25J1/00 takes precedence; programme-controlled manipulators B25J9/00 {; vehicle aspects B60, B62, e.g. remote-controlled steering for motor vehicles B62D1/24; control of position of vehicles G05D1/00}) · CPC title

• Y10S901/09

  Closed loop, sensor feedback controls arm movement · CPC title

• B25J5/007

  mounted on wheels · CPC title

• B25J5/005

  mounted on endless tracks or belts · CPC title

• B25J9/1697 Primary

  Vision controlled systems · CPC title