Distributed knowledge base method for vehicular localization and work-site management

• US 9,235,214 B2
• Filed: 09/11/2008
• Issued: 01/12/2016
• Est. Priority Date: 09/11/2008
• Status: Active Grant

• Current Assignee: Deere & Company
• Original Assignee: Deere & Company
• Inventors: Anderson, Noel Wayne
• Primary Examiner(s): Starks, Wilbert L
• Application Number:
  US12/208,782
• Publication Number:
  US 20100063954A1
• Time in Patent Office: 2,679 Days
• Field of Search: 706/50, 706/45
• US Class Current: 1/1
• CPC Class Codes: G05D 1/0088 characterized by the autonomous decision making process, e.g. artificial intelligence, predefined behaviours using knowledge based models G06N5/00
• CPC Class Codes: G05D 1/0221 involving a learning process
• CPC Class Codes: G05D 1/0274 using mapping information stored in a memory device navigation using map-matching G01C21/30
• CPC Class Codes: G06N 3/008 based on physical entities controlled by simulated intelligence so as to replicate intelligent life forms, e.g. based on robots replicating pets or humans in their appearance or behaviour

Abstract

The illustrative embodiments provide a method for controlling a vehicle. In an illustrative embodiment, a dynamic condition is identified and the vehicle is controlled using a knowledge base comprising a fixed knowledge base and a learned knowledge base.

Claims

1. A method for controlling a vehicle, the method comprising:

identifying a dynamic condition; and

responsive to identifying the dynamic condition, dynamically controlling the vehicle in real-time using a knowledge base comprising a fixed knowledge base and a learned knowledge base, wherein the fixed knowledge base comprises static information about an operating environment of the vehicle and the learned knowledge base comprises knowledge learned as the vehicle spends time in a specific work area, wherein the fixed knowledge base further comprises an online knowledge base and wherein the identifying step comprises;

receiving sensor data from a plurality of sensors for the vehicle; and

applying weights to the sensor data using the online knowledge base to form weighted sensor data that prioritizes the sensor data received from at least some of the plurality of sensors, wherein the online knowledge base dynamically provides information to a machine control process of the vehicle using a communication unit to remotely access data.

2. The method of claim 1, further comprising:

selecting a portion of the weighted sensor data to form selected sensor data; and

performing localization using the selected sensor data.

3. The method of claim 2, wherein the step of performing localization using the selected sensor data comprises:

identifying a set of objects in an environment around the vehicle using the selected sensor data; and

identifying a location of the vehicle using the location of the set of objects.

4. The method of claim 3, wherein the step of identifying the set of objects in the environment around the vehicle using the selected sensor data comprises:

identifying the set of objects using the sensor data and the knowledge base.

5. The method of claim 3, wherein the knowledge base specifies the set of objects based on the environment.

6. The method of claim 5, wherein the knowledge base specifies identifying at least one of a curb and driveway if the season is summer, and wherein the knowledge base specifies identifying at least one of a tree trunk and a tree branch if the season is winter.

7. The method of claim 5, wherein the knowledge base specifies identifying at least one of a fairway and a golfing green if the season is summer, and wherein the knowledge base specifies identifying at least one of a tree trunk and a tree branch if the season is winter.

8. The method of claim 1, further comprising:

selectively updating the learned knowledge base using the dynamic conditions.

9. The method of claim 8, wherein the selectively updating step comprises:

detecting changes to a set of objects in the environment using information about the set of objects maintained in the learned knowledge base; and

updating the learned knowledge base with the changes.

10. The method of claim 9, wherein the set of objects comprises at least one of a tree, a street, a yard, and a parked car.

11. The method of claim 1, wherein the fixed knowledge base comprises an a priori knowledge base and an environmental knowledge base, wherein the environmental knowledge base comprises environmental data.

12. A method for controlling a vehicle, the method comprising:

identifying a dynamic condition; and

responsive to identifying the dynamic condition, dynamically controlling the vehicle in real-time using a knowledge base comprising a fixed knowledge base and a learned knowledge base, wherein the fixed knowledge base comprises static information about an operating environment of the vehicle and the learned knowledge base comprises knowledge learned as the vehicle spends time in a specific work area, wherein the step of identifying the dynamic condition comprises;

identifying a set of differences between an environment as detected by a set of sensors for the vehicle and a description of the environment in the fixed knowledge base.

13. The method of claim 12 further comprising:

updating the learned knowledge base with the set of differences.

14. The method of claim 12, wherein the set of differences comprises at least one of a change in an object in the environment, a new object in the environment, and a missing object from the environment.

15. The method of claim 12, further comprising:

sending the set of differences to a server based on a policy, where the set of differences are used to update the fixed knowledge base.

16. The method of claim 15, wherein the policy is at least one of supervised and unsupervised learning, wherein supervised learning requires user input to update the fixed knowledge base, and wherein unsupervised learning updates the fixed knowledge base without user input.