Automated inspection and verification of electric motor weld quality
A method of inspecting an electric motor includes scanning an electric motor stator winding with a 2D or 3D camera, acquiring one or more images of a plurality welds between adjacent electrical wires forming the stator winding using the 2D camera, analyzing the one or more acquired images with at least one neural network such that the neural network determines if at least one of the plurality of welds has a weld defect. The at least one neural network is trained and distinguishes between surface discoloration on a surface of the welds and defect discoloration resulting from contamination during welding. Also, the method inspects over 150 welds per electric motor stator winding moving along an assembly line.
1. A method of inspecting an electric motor, the method comprising:
scanning an electric motor stator with a stator winding with a camera;
acquiring images of a plurality of welds between adjacent electrical wires forming the stator winding using the camera; and
analyzing the acquired images with at least one neural network, wherein the neural network determines if at least one of the plurality of welds has one of a weld defect, a surface discoloration, and a defect discoloration.
2. The method according to claim 1 , wherein the camera is a 2D area camera.
3. The method according to claim 1 , wherein the camera is a 2D line scan camera.
4. The method according to claim 1 further comprising transporting the electric motor stator across a scanning area of the camera such that a 2D camera scans and acquires the images of the plurality of welds as the electric motor stator moves relative to the 2D camera.
5. The method according to claim 4 , wherein the electric motor stator is transported across the scanning area of the 2D camera with a conveyor.
6. The method according to claim 1 further comprising:
transporting a plurality of electric motors across a scanning area of a 2D camera;
scanning and acquiring images of the plurality welds between adjacent electrical wires for each of the plurality of electric motors;
analyzing each of the acquired images with the at least one neural network; and
determining and distinguishing between surface discoloration and weld defects for each of the plurality of welds for each of the plurality of electric motors.
7. The method according to claim 6 , wherein the plurality of welds for each of the plurality of electric motors is more than 100 welds.
8. The method according to claim 7 , wherein each of the adjacent electrical wires for each of the plurality of electric motors is laser welded to form the plurality of welds.
9. The method according to claim 1 , wherein the weld defects comprise at least one of voids, pits, cracks, weld size too small, weld size too large, and enamel contamination.
10. The method according to claim 1 further comprising:
scanning and acquiring an image of at least one weld between an electrical wire and a connector using a 2D camera;
analyzing the acquired the image with the at least one neural network; and
determining and distinguishing between surface discoloration and weld defects for the weld.
11. The method according to claim 1 , wherein the adjacent electrical wires comprise adjacent hairpin segments.
12. The method according to claim 11 , wherein the adjacent hairpin segments are polymer coated with de-coated ends.
13. The method according to claim 12 , wherein the at least one neural network determines if any of the plurality of welds have a polymer contamination weld defect.
14. The method according to claim 1 further comprising generating an alert and an identification when a weld defect is determined for at least one of the plurality of welds, wherein the identification comprises a position of the weld defect on the stator winding.
15. A method of automatically inspecting a plurality of electric motors on an assembly line, the method comprising:
transporting electric motor stators on a conveyor through an inspection station comprising a 2D camera, wherein each of the electric motor stators has a stator winding formed from a plurality of hairpin wire segments joined together with a plurality of welds;
scanning and acquiring images of the plurality welds with the 2D camera;
analyzing the acquired images with at least one neural network configured to determine and distinguish between surface discoloration and weld defects on a surface of a weld; and
generating an alert when a weld defect is determined for at least one of the plurality of welds.
16. The method according to claim 15 , wherein the weld defects are selected from the group consisting of voids, pits, cracks, weld size too small, weld size too large, and enamel contamination.
17. The method according to claim 15 , wherein the alert comprises an identification of a weld defect type and a position of the weld defect on the stator winding.
18. A method of automatically inspecting a plurality of electric motors on an assembly line, the method comprising:
transporting electric motor stators on a conveyor through an inspection station across a scanning area of a 2D camera, wherein each of the electric motor stators has a plurality of hairpin wire segments with a plurality of welds joining adjacent hairpin wire segments;
scanning and acquiring images of the plurality of welds using the 2D camera; and
analyzing the acquired images with at least one neural network configured to determine and distinguish between surface discoloration and weld defects on a surface of a weld;
determining via the at least one neural network if any of the plurality of welds have a weld defect;
generating an alert when a weld defect is identified on at least one of the plurality of welds; and
generating an identification for each of the plurality of welds having the weld defect.
19. The method according to claim 18 further comprising training the at least one neural network with an identified weld defect.