Using planar sensors for pallet detection

Patent metadata

Field	Value
Publication number	US-9715232-B1
Application number	US-201615268984-A
Country	US
Kind code	B1
Filing date	Sep 19, 2016
Priority date	Sep 19, 2016
Publication date	Jul 25, 2017
Grant date	Jul 25, 2017

Abstract

Example implementations may relate to a mobile robotic device that is operable to detect pallets using a distance sensor. According to these implementations, the robotic device causes the distance sensor to scan a horizontal coverage plane in an environment of the robotic device. Then, the robotic device receives from the distance sensor, sensor data indicative of the horizontal coverage plane. The robotic device compares the sensor data to a pallet identification signature. Based on the comparison, the robotic device detects a pallet located in the environment. Further, based on the sensor data, the robotic device determines a location and an orientation of the detected pallet.

First claim

The invention claimed is:

1. A method comprising:
causing, by a control system, a distance sensor coupled to a robotic device to scan a horizontal coverage plane in an environment of the robotic device;
receiving, by the control system from the distance sensor, first sensor data indicative of the horizontal coverage plane;
comparing, by the control system, the first sensor data to a pallet identification signature, wherein the pallet identification signature is indicative of at least two dimensions of a support member of a pallet type;
based on the comparison, detecting, by the control system, a pallet located in the environment; and
based on the first sensor data, determining, by the control system, a location and an orientation of the pallet.

2. The method of claim 1 , wherein the first sensor data comprises a plurality of curves in a two-dimensional representation of the environment, wherein each of one or more points along each of the curves is indicative of a horizontal distance to a surface in the environment.

3. The method of claim 2 , wherein comparing, by the control system, the first sensor data to the pallet identification signature comprises:
detecting a first corner pattern in a first curve using the first sensor data, wherein a corner pattern comprises a substantially 90 degree change in a slope of a line, and wherein the first corner pattern is indicative of a first corner of a first candidate support member;
detecting a first discontinuity and a second discontinuity in the first curve using the first sensor data, wherein the first discontinuity and the second discontinuity are indicative of a second corner and a third corner of the first candidate support member respectively;
based on respective locations of each of the first corner, the second corner, and the third corner, determining (i) a width of the first candidate support member, and (ii) a length of the first candidate support member; and
comparing the length and the width to the at least two dimensions of the support member indicated by the pallet signature.

4. The method of claim 3 , wherein detecting, by the control system, the pallet located in the environment comprises:
determining that the length and the width correspond to the at least two dimensions of the support member of the pallet type indicated by the pallet signature.

5. The method of claim 3 , wherein comparing, by the control system, the first sensor data to the pallet identification signature further comprises:
determining a second corner pattern in a second curve, wherein the second corner pattern is indicative of a first corner of a second candidate support member;
analyzing the first sensor data to determine a first discontinuity and a second discontinuity in the second curve, wherein the first discontinuity and the second discontinuity in the second curve are indicative of a second corner and a third corner of the second candidate support member respectively;
based on respective locations of each of the first corner, the second corner, and the third corner of the second candidate support member, determining a location of the second support member; and
determining a distance between the second candidate support member and the first candidate support member.

6. The method of claim 5 , wherein based on the comparison, detecting, by the control system, a pallet located in the environment comprises:
determining that the distance between the first candidate support member and the second candidate support member corresponds to a pallet pocket width of the pallet type; and
determining that the first candidate support member and the second candidate support member are support members of the pallet.

7. The method of claim 1 , wherein detecting a pallet located in the environment further comprises determining the pallet type of the detected pallet, wherein the pallet type is indicative of: (i) a number of the at least one physical support feature of the detected pallet, (ii) dimensions of the detected pallet, and (iii) a number and location of tine openings of the detected pallet.

8. The method of claim 7 , wherein the control system determining a location and orientation of the detected pallet in the environment is further based on the pallet type of the detected pallet, and wherein determining the orientation of the pallet comprises estimating a six degree of freedom (DOF) pose of the detected pallet.

9. The method of claim 1 , wherein the robotic device is a forklift robotic device comprising two tines, wherein the horizontal coverage plane is parallel to the two tines, and wherein the method further comprises:
operating the robotic device to insert each of the two tines into a respective tine pocket;
after inserting the tines, receiving from the distance sensor, second sensor data comprising a plurality of curves in a two-dimensional map of the environment, wherein each of one or more points along each curve is indicative of a respective horizontal distance to one or more surfaces in the environment;
analyzing the second sensor data to detect a first corner pattern in a first curve, wherein the first corner pattern comprises a substantially 90 degree change in a slope of the first curve, and wherein the first corner pattern is indicative of a front corner of the detected pallet;
analyzing the second sensor data to detect an inflection point in the first curve, wherein the inflection point is indicative of a back corner of the detected pallet; and
based on a location of the front corner and a location of the back corner of the support member, determining a positioning of the tines with respect to the pallet.

10. The method of claim 9 , the method further comprising:
based on the positioning of tines with respect to the pallet, determining an adjustment to center the tines within the tine pockets; and
responsively adjusting the tines.

11. A robotic device comprising:
a distance sensor coupled to the robotic device;
two tines; and
a control system operable to:
cause the distance sensor coupled to a robotic device to scan a horizontal coverage plane in an environment of the robotic device;
receive from the planar distance sensor, first sensor data indicative of the horizontal coverage plane;
compare the first sensor data to a pallet identification signature, wherein the pallet identification signature is indicative of at least two dimensions of a support member of a pallet type;
based on the comparison, detect a pallet located in the environment; and
based on the first sensor data, determine a location and an orientation of the detected pallet.

12. The robotic device of claim 11 , further comprising a second planar sensor, and wherein each of the two planar sensors is mounted on a respective tine of the two tines.

13. The robotic device of claim 11 , wherein the distance sensor is a light detection and ranging (LIDAR) sensor, and wherein the method further comprises causing the LIDAR sensor to scan the horizontal coverage plane by rotating about an axis perpendicular to a plane of motion of the robotic device.

14. The robotic device of claim 13 , wherein the LIDAR sensor rotates 270 degrees about the axis, and wherein the horizontal coverage plane comprises a 270-degree horizontal field-of-view (FOV) around the robotic device.

15. The robotic device of claim 11 , further comprising:
an adjustable mounting plate coupled to the robotic device, wherein the planar sensor is mounted onto the adjustable mounting plate, and wherein the mounting plate is configured to adjust the positioning of the planar sensor such that the horizontal coverage plane i

Assignees

• X Dev Llc

Inventors

• Fischer Mark
• Zevenbergen John

Classifications

• G01S17/08

  for measuring distance only (indirect measurement G01S17/46; active triangulation systems G01S17/48) · CPC title

• G05D1/0231 Primary

  using optical position detecting means (position-fixing by using electromagnetic waves other than radio waves, e.g. optical position detecting means G01S5/16) · CPC title

• Y10S901/01

  Mobile robot · CPC title

• G05D2201/0216

  Physics · mapped topic

• B66F9/063

  Automatically guided · CPC title