Granted Patent B2

US 12,232,687 · App. 17/735,660 · Granted Feb 25, 2025

Inventors: Soonhyuk Hong (Suwon-si, KR); Minjeong Kang (Suwon-si, KR); Sejin Kwak (Suwon-si, KR)

Assignee: Samsung Electronics Co., Ltd.

A47L9/2852A47L9/009A47L9/2826G05D1/0212G05D1/0231A47L2201/04A47L2201/06G05D1/0038

Abstract

Claims (51)

1. A method, performed by a cleaning robot, of performing a task, the method comprising:

capturing an image of an object in a vicinity of the cleaning robot;

determining a type of the object, by applying the captured image to at least one trained artificial intelligence (AI) model; and

determining driving path of the cleaning robot, according to the type of the object,

wherein the driving path includes:

performing moving in close contact with the object when the type of the object is determined to be a fixed object or a hard to move object,

performing moving away from the object by a certain distance when the type of the object is determined to be a fragile object, and

performing interaction toward the object when the type of the object is determined to be at least one of a person, a body part of a person, and animal.

2. The method of claim 1 ,

wherein the object is determined as a first object, and

wherein the driving path of the cleaning robot is determined as close avoidance driving in which a first distance is spaced apart from the first object.

3. The method of claim 2 ,

wherein the first object has a property of being fixed in the area or difficult to move, and

wherein the first distance is less than 1 cm.

4. The method of claim 2 , further comprising:

measuring distance information between the first object and the cleaning robot in real time,

wherein the driving path of the cleaning robot is determined such that the cleaning robot is separated from the first object by the first distance.

5. The method of claim 1 ,

wherein the object is determined as a second object, and

wherein the driving path of the cleaning robot is determined as complete avoidance driving in which a second distance is spaced apart from the second object.

6. The method of claim 5 ,

wherein the second object has a fragile property or a contamination possibility property, and

wherein the second distance is 1 cm or more and 15 cm or less.

7. The method of claim 5 , further comprising:

measuring distance information between the second object and the cleaning robot in real time,

wherein the driving path of the cleaning robot is determined such that the cleaning robot is separated from the second object by the second distance.

8. The method of claim 5 , further comprising:

measuring distance information between the second object and the cleaning robot in real time,

wherein a virtual wall is set in a second area separated from a second object by the second distance, and the cleaning robot drives in close contact with the virtual wall.

9. The method of claim 1 ,

wherein the object is determined as a second object, and

wherein the cleaning robot performs an interaction toward a third object.

10. The method of claim 9 , wherein the third object has a property of reacting to the interaction.

11. The method of claim 10 , further comprising:

measuring a position of the third object moved in response to the interaction and distance information between the third object and the cleaning robot in real time,

wherein the driving path of the cleaning robot is determined such that the cleaning robot is separated from the second object by a predetermined distance.

12. The method of claim 1 , wherein the determining of the task to be performed by the cleaning robot comprises:

obtaining identification information of the object by applying the captured image to a trained first AI model; and

determining the driving path of the cleaning robot by applying the obtained identification information of the object to a trained second AI model.

13. A cleaning robot comprising:

a camera;

a driver; and

at least one processor configured to:

control the camera to capture an image of an object in a vicinity of the cleaning robot,

determine a type of the object, by applying the captured image to at least one trained artificial intelligence (AI) model,

determine driving path of the cleaning robot, according to the type of the object, and

control the driver to drive the cleaning robot towards the object according to the driving path of the cleaning robot,

wherein the driving path includes:

performing moving in close contact with the object when the type of the object is determined to be a fixed object or a hard to move object,

performing moving away from the object by a certain distance when the type of the object is determined to be a fragile object, and

performing interaction toward the object when the type of the object is determined to be at least one of a person, a body part of a person, and animal.

Priority Claims (1)

Continuity (2)

References Cited (85)