RMAL home > Research > Automated Inspection
MMRI | MSL | CAPPA-D | Robotics

 

Automated Inspection Using Machine Vision 

1.  Introduction:

We have developed several automated inspection systems for industry, as follows:

2.  High-Speed Inspection Machine for Automotive Parts

We have developed a machine we developed for a local company that inspects the bearing surfaces on automotive rocker arms using machine vision; and then sorts them into reject/accept bins. Subtle surface defects can be detected, and the machine can inspect 60 parts per minute. The machine is now running in-production at the customer's facility. Video of the machine in operation (1.3 meg AVI file using DIVX compression, the CODEC is available here or from www.divx.com).

The rocker arm inspection machine.

Detection of a defect (raised metal) on the dimple-shaped bearing surface.

3.  Automated Inspection of Subtle Surface Defects

We have developed a system incorporating custom lighting and image processing techniques which is capable of detecting very subtle defects on the surface of products.  These defects can even be difficult for a person to see under ambient lighting.  Examples are shown in Figures 1-4.  As these examples demonstrate, the system can successfully detect subtle defects on flat and curved surfaces. This system was completed in Sept. 2000 and has been transferred to the end user’s facility.

Figure 1:  Product with a subtle dent on its curved edge.  A: Under ambient lighting.  B: Subtle dent located by the automated inspection system.


Figure 2:  Product with a subtle dent on its flat face.  A: Under ambient lighting.  B: Subtle dent located by automated inspection system.

Figure 3:  Product with a dust particle near its right edge.  A: Under ambient lighting.  B: Dust particle located by automated inspection system.

Figure 4:  Product with a subtle scratch.  A: Under ambient lighting.  B: Subtle scratch located by the automated inspection system.


4. Resistance Welding Machine Incorporating Automated Inspection

4.1 Overview

This machine automated the resistance spot welding of two small wires, a job previously done manually using a microscope and tweezers.  Part handling, weld power monitoring, and automated inspection systems were designed and built (including hardware and software).  The machine was PC controlled with a custom user interface.

4.2 Automated Visual Inspection of the Two Welds

The two welds are shown in Figure 5.  The first weld, shown at the left, involves the connection of a 0.4 mm diameter niobium wire to a block, and was termed the “block-weld”.  The second weld, shown on the right, involves the connection of two 0.4 mm dia. wires together to form a T shape, and was termed the “T-weld”. A close-up of a sample T-weld is shown in Figure 6.  The welds were inspected for several dimensional specifications using machine vision.  The primary specifications measured were the penetration of the wire into the block with the block-weld, and the penetration of the two wires into each other with the T-weld.  This penetration is also known as embedment.  Two sample inspection results for the T-weld are shown in Figure 7.  Some of the dimensions have been overlayed on the original camera image. The T-weld inspection system involved custom lighting and image processing software development.  The block-weld inspection system employed the standard laser striping technique with custom image processing software.  A sample result with the laser stripe shown is given in Figure 8.

Figure 5.  The two resistance welds the machine was built to automate.

Figure 6.  Close-up of a T-weld.

Figure 7.  Two Sample automated inspection results for the T-weld.  Some the measurements have been overlayed on the original image. (Please Note:  These are photos taken of the computer screen at the end-user's facility).

Figure 8.  An image taken during the inspection of a block-weld showing the laser stripe.  (Photo of the computer screen at the end-user's facility).

4.3 System Construction, Transfer and In-Production Use

The system was developed at our lab in McMaster University and transferred to the end-user’s facility upon completion.  The system is shown during the development phase in Figure 9.  The system is shown operating in-production at the end user’s facility in Figure 10.  The system was completed in July 1994 and ran on-line at the end user’s facility until May 2000 (ending only because the product being welded was discontinued).

Figure 9.  The resistance welding machine during development at McMaster.  The motors for the conveyor system can be seen near the bottom left of the left image.

Figure 10.  The completed resistance welding machine operating in-production inside the end user's clean room.


5.  Laser Welding Machine Incorporating Automated Inspection

5.1 Overview

This machine automated the laser welding of two formed titanium parts (0.5 mm wall thickness). A 4-axis servo controlled fixturing unit (for manipulating the parts relative to the stationary laser), and an automated inspection system (hardware and software) were developed. The machine was PC controlled with a custom user interface. An existing laser was re-used, again controlled by the PC.

4.2 Automated Pre and Post Weld Visual Inspection

Since the laser welding involves no filler metal, the gap between the two parts must be within a small tolerance to provide a continuous, void-free weld.  The alignment along the seam was measured in the out-of-plane dimension using the standard laser striping technique while custom image processing was used to measure the in-plane gap between the parts.  Sample images are shown in Figure 11.   If the alignment of the parts was found to be acceptable by the inspection system the automated welding was then performed.  A post-weld machine vision inspection was then performed (images unavailable).

Figure 12.  Measurement of the in-plane gap between the parts prior to welding.  Left: Original camera image.  Right: During image processing.  (Please Note: These are photos taken of the computer screen at the end-user's facility).

5.3 System Construction, Transfer and In-Production Use

The system was developed at McMaster and transferred to the end-user’s facility upon completion.  The system is shown during the development phase in Figure 13, with the 4-axis servo controlled fixturing unit at the top, right.  The system is shown at the end user’s facility in Figure 14.  The system was completed in May 1994 and continues to run in-production at the end user’s facility.

Figure 13 (at left):  Laser welding machine during development at McMaster University.

Figure 14 (at right):  The automated laser welding machine installed at the end user's facility.


For Further Information Please Contact:

Dr. Gary M. Bone
Dept. of Mechanical Engineering
McMaster University
Hamilton, Ontario, CANADA, L8S 4L7.
Phone: (905) 525-9140 ext. 27591
Fax: (905) 572-7944
e-mail: gary at mcmaster.ca

Researchers:

Dr. Gary Bone, Dr. David Capson, Peter Bender, Lucian Balan, Yan Li, Robert van Varseveld.
Contact Us | Legal |Privacy | Our Site