Method and system for the directed control of robotic assets

• US 9,969,081 B2
• Filed: 07/26/2013
• Issued: 05/15/2018
• Est. Priority Date: 07/27/2012
• Status: Active Grant

• Current Assignee: Robotic Research OPCO LLC
• Original Assignee: Karl Murphy
• Inventors: Lacaze, Alberto Daniel, Murphy, Karl, Legowik, Steven Anthony, Rayej, Shima
• Primary Examiner(s): Tran, Khoi H
• Assistant Examiner(s): Peche, Jorge O
• Application Number:
  US13/952,542
• Publication Number:
  US 20140058563A1
• Time in Patent Office: 1,754 Days
• Field of Search: 700248, 700253, 700258, 701409, 701410, 701411, 701412, 701446, 701454
• CPC Class Codes: B25J 9/162 Mobile manipulator, movable base with manipulator arm mounted on it
• CPC Class Codes: G05D 1/0274 using mapping information stored in a memory device navigation using map-matching G01C21/30
• CPC Class Codes: G05D 1/0278 using satellite positioning signals, e.g. GPS
• CPC Class Codes: G05D 1/0295 by at least one leading vehicle of the fleet

Abstract

Static multi-registration performs a delayed convoy mission from a recorded path. The data recorded for delayed convoy includes both positions traversed by the vehicle and the obstacles sensed as it drove along the path and generates a new path file in this format that is used by the vehicle to follow the desired route. The data is processed before it is passed to the vehicle to be followed. Paths are processed to determine where they cross each other. These intersection points and the path data are used to create an interconnected graph of path segments. A multi-registration planner uses that information on the length and directionality of the path segments to compute the best route between two intersections. The route generated by the planner guides the merging of position and obstacle data from several recorded paths into a single record used by the vehicle to follow the desired route.

Claims

1. A method for controlling robotic assets comprising the steps of:

providing a computer server;

providing a database, stored on the computer server, of previously collected features collected by a robotic asset;

providing a database, stored on the computer server, containing traversable trajectories defining a road network;

using a network of features, in real time, to localize a vehicle;

sensing features of a terrain or an environment by the vehicle where features are defined as any physical attribute, that can be sensed with onboard sensors;

creating one or more trajectories collected by one or more assets into a single trajectory that is followed by the vehicle;

using a vehicle navigation system to control the vehicle'"'"'s movement based on the created trajectory;

generating an interconnected graph, where each segment of the graph can have a different frame and where each segment is connected to other segments by matching features that are sensed by the vehicles in the intersections;

searching through the generated graph to compute the best route;

producing a set of concatenated edges from the generated graph that defines the best route or optimal trajectory;

storing features collected by multiple assets in the data base; and

using vehicles that are in a GPS-denied area, or in an areas where GPS does not provide sufficient accuracy for driving.

2. The method of claim 1, wherein the feature matching between the feature server and the vehicle is done using mapettes.

3. The method of claim 2, wherein

the features within a sensor'"'"'s range are sent to the vehicle in conjunction with the trajectory of the original asset;

and

sensor-based registration calculates and corrects for relative drift.

4. The method of claim 3, further comprising a matching algorithm comparing the features on the database and the features sensed at that moment in time to determine an error between any two vehicles;

and

wherein the first vehicle is the one that collected an original features and the second vehicle is the one that is following the path.

5. The method of claim 4, further comprising a filtering algorithm correcting the navigation solution on the vehicle given a delta pose error.

6. The method of claim 5, further comprising the step of

utilizing features collected by multiple assets into a mapette that can be used by a Kalman filtering algorithm to correct a navigation solution.

7. The method of claim 2, wherein the features within a sensor'"'"'s range of the vehicle are stored on the vehicle itself.

8. The method of claim 7, further comprising a matching algorithm comparing the features on the database and the features sensed at that moment in time to create a delta pose error.

9. The method of claim 8, further comprising a filtering algorithm correcting the navigation solution on the vehicle given the delta pose error.

10. The method of claim 9, further comprising a control algorithm steering the vehicle from its current pose to the trajectory provided from the database and the corrected navigation solution.

11. The method of claim 1, further comprising the step of

stringing or concatenating multiple trajectories collected by different assets into a single trajectory that can then be followed by the vehicle.

12. The method of claim 11, further comprising the step of

utilizing features collected by multiple assets into a coherent map that can be used by the filtering algorithm to correct a navigation solution.

13. The method of claim 1, further comprising the step of

utilizing a search algorithm to calculate a best route through the road network utilizing segments collected by different assets.

14. The method of claim 1, wherein features collected by the vehicle are used to complement the database of features.