Robotic self programming visual inspection
A robotic self-learning visual inspection method includes determining if a fixture on a component is known by searching a database of known fixtures. If the fixture is unknown, a robotic self-programming visually learning process is performed that includes determining one or more features of the fixture and providing information via a controller about the one or more features in the database such that the fixture becomes known. When the fixture is known, a robotic self-programming visual inspection process is performed that includes determining if the one or more features each pass an inspection based on predetermined criteria. A robotic self-programming visual inspection system includes a robot having one or more arms each adapted for attaching one or more instruments and tools. The instruments and tools are adapted for performing visual inspection processes.
1. A robotic self-learning visual inspection method, comprising:
providing a robot having one or more arms each adapted for attaching to one or more of a plurality of tools;
providing a controller configured to instruct the robot to perform the steps of:
imaging a component using a long-range camera;
searching images of the component for a fiducial marking to identify a fixture;
scanning the fixture with a touch probe along (1) a longitudinal direction of the fixture at different transverse positions, and (2) a transverse direction, perpendicular to the longitudinal direction, at different longitudinal positions; and
determining widths of sections of the fixture between scans with the touch probe via a distance sensor;
generating a computational model of the fixture based on scanning the fixture and determining the widths of sections;
retrieving data from a database for the fixture;
identifying one or more features on the fixture based on the computational model and data from the database;
capturing close-up images using a short-range camera for collecting detailed information of the one or more features on the fixture;
comparing the data from the database with data from the close-up images; and
determining whether each of the one or more features passes inspection based on predetermined criteria.
2. The robotic self-learning visual inspection method of claim 1 , further comprising prompting a user to scan a barcode attached to the fixture.
3. The robotic self-learning visual inspection method of claim 1 , wherein generating the computational model further comprises:
generating a plurality of waypoints;
building an itinerary to provide safe travel for movement of the robot near the fixture based on the plurality of waypoints; and
validating the itinerary to ensure that the plurality of waypoints are reachable by at least one arm of the robot.
4. The robotic self-learning visual inspection method of claim 1 , wherein information extracted from images of the fixture is projected onto the computational model to provide three-dimensional coordinate information of the feature.
5. The robotic self-learning visual inspection method of claim 1 , further comprising uploading feature information to the database such that a previously unknown fixture becomes known.
6. The robotic self-learning visual inspection method of claim 1 , further comprising examining a feature, wherein the controller processes images from the short-range camera for finding measured locations in the images.
7. The robotic self-learning visual inspection method of claim 1 , further comprising:
pausing the method when a failure is identified by the controller;
moving a laser via the one or more arms of the robot; and
illuminating a failure location directly on the fixture with the laser.
8. The robotic self-learning visual inspection method of claim 1 , wherein the plurality of tools are selected from a long-range camera, a short-range camera, a barcode scanner, a distance sensor, a touch probe, a light, and a laser.
9. The robotic self-learning visual inspection method of claim 1 , wherein the fixture comprises a bond, a fastener, a component edge, or a component corner.
10. The robotic self-learning visual inspection method of claim 1 , wherein the feature comprises a slider, a header-board, a bolt, a bolt slot, a fastener edge, a layup pattern, a seam, a bond, or an overlap.
11. The robotic self-learning visual inspection method of claim 1 , comprising determining that the features is out of position based on the predetermined criteria, wherein the predetermined criteria comprises a tolerance limit.
12. The robotic self-learning visual inspection method of claim 1 , comprising repeating steps provided by the controller for a plurality of features on the fixture.
13. The robotic self-learning visual inspection method of claim 12 , comprising repeating the steps provided by the controller for a plurality of fixtures on a component.
14. The robotic self-learning visual inspection method of claim 13 , comprising providing a pass status for the component when the predetermined criteria have been met for the plurality of fixtures on each of the plurality of features.
15. The robotic self-learning visual inspection method of claim 1 , comprising moving the robot along a track mounted to a floor for position the one or more arms with respect to the fixture.
16. The robotic self-learning visual inspection method of claim 1 , comprising receiving user input about the fixture via a user interface.
17. The robotic self-learning visual inspection method of claim 1 , comprising preplacing a fiducial marking on a component, the fiducial marking being configured such that the controller may identify the fixture using one or more of the plurality of tools.