Human-Guided Mapping Method for Mobile Robot

• US 20090198375A1
• Filed: 10/18/2006
• Published: 08/06/2009
• Est. Priority Date: 10/18/2006
• Status: Active Grant

• Current Assignee: Yutaka Kanayama
• Original Assignee: Yutaka Kanayama
• Inventors: Kanayama, Yutaka
• Granted Patent: US 8,068,935 B2
• Time in Patent Office: Days
• US Class Current: 700/245
• CPC Class Codes: G05B 2219/40446 Graph based
• CPC Class Codes: G05B 2219/40506 Self motion topology knowledge, configuration mapping
• CPC Class Codes: G05B 2219/40508 Fuzzy identification of motion plans executed by operator
• CPC Class Codes: G05D 1/0221 involving a learning process

Abstract

A method of mapping an operation area by a team of a human and a mobile robot 200 includes the steps of defining a graph representing the area by a human 201, guiding the robot by human along an edge 203, stopping at a vertex in the graph by the team 203, creating a vertex record if stopped at a new vertex 205, localizing the robot and vertices if stopped at an existing vertex 206, creating an edge record if finished a new edge 208, and outputting an area'"'"'s map including a set of vertex records and a set of edge record by the robot 210. The robot'"'"'s human-tracking step 203 includes the steps of obtaining a 2-DOF motion command from sensors that detect the human'"'"'s action and executing a 2-DOF motion based on the motion command.

Claims

1. A method of enabling a mobile robot to track a human in an operation area comprising the steps of:

obtaining a two-variable motion command from sensors that detect the human'"'"'s action; and

executing a motion of the robot with a translation speed and a rotation speed based on the motion command, while keeping the robot'"'"'s translation direction the same as the robot'"'"'s body direction.

2. The method of claim 1, wherein the translation speed in the executing step is restricted to non-negative.

3. The method of claim 1, wherein the translation speed in the executing step is restricted to non-positive.

4. A method of enabling a team of a human and a mobile robot to map an operation area comprising the steps of:

defining a graph that is embedded in the operation area by the human;

traversing an edge in the graph while the robot tracks the human using the method of claim 2;

extracting geometrical features of the both-side objects by the robot while traversing an edge in the graph;

stopping at a vertex in the graph at the end of edge traversing;

telling the robot the number of the vertex by the human when the team stops at a vertex;

creating a vertex record that includes a vertex number, an anchor, a human-provided name, and a vertex position by the robot when the team stops at a new vertex;

localizing the robot and vertices by the robot when the team stops at an existing vertex;

creating an edge record that includes a pair of vertex numbers, a distance, geometrical features of the both-side objects by the robot when the team finishes traversing of a new edge; and

outputting the set of the created vertex records and the set of the created edge records as a map of the operation area by the robot at the end of the mapping session.

5. The method of claim 1, wherein further comprising the steps of:

creating a motion record that includes a translation speed and a rotation speed at each motion-control interval; and

outputting the set of the created motion records at the end of the human-tracking motion.

6. The method of claim 5, wherein further comprising the step of reproducing the human-tracking motion using the set of the motion records.

7. The method of claim 5, wherein further comprising the step of reproducing the human-tracking motion in a reverse order using the set of the motion records.