NI Compact Vision System Helps Achieve Six-Sigma Repeatability Standards in Inspection of Automobile Spark Plugs

Anand Krishnan, Soliton Technologies Pvt. Ltd.

"The result is an inexpensive, rapidly developed inspection system with the highest commitment to quality."

- Anand Krishnan, Soliton Technologies Pvt. Ltd.

The Challenge:

Developing a robust and reliable automated optical inspection system for automobile spark plugs while meeting exacting standards of six-sigma gage reproducibility, repeatability, accuracy, and saving floor space.

The Solution:

Architecting a rugged, reliable, yet inexpensive shop-floor inspection solution based on the National Instruments Compact Vision System, a Sony IEEE 1394 digital camera and flexible, modular software written on an NI LabVIEW Real-Time and NI Vision Development Module platform.

Author(s):

Anand Krishnan - Soliton Technologies Pvt. Ltd.
S. Vasanth Prabhu - Soliton Technologies Pvt. Ltd.
Karthik Nanjappan - Soliton Technologies Pvt. Ltd.
Anand Prasad Chinnaswamy - Soliton Technologies Pvt. Ltd.

 

 

Proper sparking distance is critical in the production of an automobile spark plug. Our customer, the leading Indian manufacturer of spark plugs, identified two key dimensional parameters that affect the functional performance of the spark plug. One is the eccentricity between the outer housing profile and the inner electrode. The other is the offset between the earth electrode and the center of the inner electrode.

 

The customer had used a manual process for measuring eccentricity and offset. They did this by finding the position of three points on each circular part and fitting the circle equation using this data. A measurement error on even one of these points could significantly skew the calculated eccentricity and offset. The lesser accuracy of this method forced the customer to narrow their tolerance band and this decreased yield. To ensure reliable quality control, faster inspection, and good return of investment, the customer asked us to build a vision-based dimensioning system for the spark plug production line.

 

The customer had the following system requirements:

  • Measure eccentricity and offset to an accuracy of 0.01mm with a six-sigma repeatability of better than 10 percent of the tolerance value.

  • Ensure that system functionality is independent of the inherent variations in plug texture from piece to piece and model to model.

  • Minimize floor space while resistant to the harsh production environment and maximize productivity by decreasing time of inspection.

  • Interface with the line PLC to remove human intervention, thus making the measurement process quick, objective, and repeatable.

  • Store records of the results for further analysis and improvement in process and yield.

 

We selected a 1280 x 960 pixel Sony FireWire digital camera with many programmable features required for this application.

 

We wrote the application software in LabVIEW Real-Time, which gave the customer, among other features, the flexibility to configure the system via TCP/IP, periodically calibrate the system using a calibration target, and transfer results data through FTP. We selected the Compact Vision System (CVS) as the platform on which this software would run, keeping in mind its real-time performance, ruggedness, small size, and our experience in using NI technology to deliver world class vision solutions. True to our expectations, the result was a zero-manual intervention system with a gage repeatability and reproducibility of less than 10 percent of the tolerance band.

 

Hardware and Software System

The inspection station consists of an IEEE 1394 Sony camera, a Nikon lens, and an Advanced Illumination Ring light to axially illuminate the top face of the spark plug. We fed the image stream into the FireWire port of an NI CVS. The spark plugs arrived on the conveyor at a speed of 1 plug per second. The digital port of the CVS communicates with the line PLC and relays used for ejection. The CVS interfaced to the factory network, so we used a standard PC running Windows to remotely configure setup parameters and store result data. To help in the calibration of the system, we built a master calibration target.

 

Because the application required repeatability of two pixels, it was important to design algorithms so the tuning parameters were in the middle of a fairly wide tolerance band. This ensured that changes in the image due to camera and imaging repeatability did not affect algorithm results. It was in this phase of algorithm prototyping that the NI Vision Assistant tool proved instrumental. It was possible to cyclically inspect many stored plug images and tabulate intermediate results using the batch processing feature, so parameter tuning became systematic and goal-oriented as opposed to a trial and error approach. While Vision Assistant proved useful for prototyping, the application needed the full versatility of LabVIEW, which NI provides in the CVS.

 

The NI Compact Vision System added value in the specific area of the circle detection algorithm. The NI circular edge detection uses a large number of points to fit the circle. Moreover, the algorithm identifies and neglects edges that are slightly off from the dominant circle, thus improving the circular fit. Another useful feature common to all of the NI edge detection algorithms is the ability to fix tradeoff between accuracy and processing time by controlling the extent of sub-sampling.

 

NI vision technology gives us the advantage of rugged, dependable, and real-time Compact Vision System performance combined with the unifying nature of LabVIEW as the software platform. The job of a vision system engineer is gradually evolving from writing and debugging thousands of lines of code to architecting creative and effective ideas for solving image processing problems because National Instruments technology is available to validate, enhance, and translate these ideas into a production ready application. This means that we can cut development time and costs. This system took us just eight man-weeks to develop and validate.

 

System Benefits

Using NI products, we realized the following system benefits:

Comparison Parameter

Earlier Manual Inspection Process

NI- Based Automated Machine Vision System

Objectivity and Repeatability

Subjective and nonrepeatable

The measurement process is repeatable and objective. The repeatability bettered six sigma standards.

Compactness and Ruggedness

Not compact, limited ruggedness

The heart of the solution is a rugged box smaller than the size of a brick.

Process Control

Task is cumbersome because data is manually tagged and captured.

Shift-wise or batch-wise, we can retrieve statistical information can be use it to improve the manufacturing process because data capture is automatic.

Yield

Because accuracy was limited, the tolerance band had to be narrowed. This decreased yield as the process would reject some borderline components that were actually acceptable.

With higher accuracy, the customer could relax tolerance, thus increasing yield.

 

Reliability and Repeatability with CVS and LabVIEW

The hallmark of a successful measurement application is its reliability and repeatability, and this goal is even more difficult in a vision-based application. Our spark plug inspection system achieved these targets, proving that both the individual hardware components and the software that ties them together are reliable and repeatable. While the rugged Compact Vision System formed the heart of the application, a flexible, well-prototyped and thoroughly tested application based on LabVIEW gave accurate and repeatable results.

 

The result is an inexpensive, rapidly developed inspection system with the highest commitment to quality. Virtual instrumentation helped seamlessly integrate the inspection system into the factory production line.

 

Author Information:

Anand Krishnan
Soliton Technologies Pvt. Ltd.
# 683 15th Cross Road 2nd Phase, JP Nagar
Bangalore 560078
India
Tel: +91 -80- 51208600
Fax: +91 -80- 51208700
kanand@solitontech.com

Figure 1. Hardware Schematic of the Spark Plug Inspection System