Featured Product
This Week in Quality Digest Live
CMSC Features
Etienne Nichols
What’s the difference?
Atul Minocha
It’s all about ROI
Ryan E. Day
September Speaker Series in review
Ryan E. Day
Realigning a cornerstone of industry
David H. Parker
Practical implications for electronic distance measurement

More Features

CMSC News
Expands simulations for electrification and aerospace design
VSL hosts special edition of show at new center in Rotterdam
Latest line touts comprehensive coverage, ease of use
Blackfly S GigE line grows
Voting for 2022–2023 term will open at CMSC 2022 starting July 25
New facility in Toronto area will showcase multiple Hexagon product lines
API division named ‘Top External Provider 2018’
Exact Metrology selected for project
Faster and more powerful than ever before

More News

Ken Vakil

Ken Vakil’s default image

CMSC

Automating Data Analysis and Report Generation of Key Characteristics Measurements

Published: Monday, August 22, 2011 - 12:41

The purpose of this article is to discuss automated analysis and report generation of key characteristics measurement data. Key characteristics (KCs) are those features of a part whose measurements must be kept to the nominal values through process control to minimize the "Taguchi Loss." KC measurements are taken using different types of metrology heads, such as laser trackers, or other types of scanners. The actual process discussed in this article requires gathering KC data using laser trackers. The metrology data are imported into software to compare the actual values with their nominal values. Thereafter, the data go through several file format transformations before the reports are created in the format required by the customer.

This article examines current manual processes and proposes an automated solution that can be implemented throughout the entire assembly line.

Key characteristics: definition

Key characteristics are features of a process, tool, part, or assembly that can have a negative impact on product performance (e.g., form, fit, and function) when it varies from its nominal dimension. To be useful, KCs must be quantifiable and measurable.

Figure 1 shows the parabolic relationship between the variable value and the cost of quality (Y axis), or Y = AX2. As the variable moves away from its nominal value, the resulting loss is referred to as the "Taguchi Loss." Both the KC and non-KC functions are also displayed.


Fig. 1: KC and non-KC function

 

KC identification process

KC identification and control has four distinct steps during the design and development stage (figure 2):

1. The KC identification process begins during an early design stage, with the flow moving down from top-level customer requirements. These requirements are converted into measurable KCs, i.e., outer mold line features such as step, gap, or fastener flushness, which are directly derived from the customer requirements.
2. A review of selected assemblies, including part location and assembly sequencing methods, is then performed.
3. In the variation and risk analysis step, KC performance is predicted for each assembly strategy.

4. Predicted performances are then compared, and risk is mitigated. Based on the variation simulation study, a final set of assembly KCs are identified and flowed down to manufacturing for monitoring and control.

 

During the production stage, KCs are monitored and feedback is sent to the engineering department to validate the design.


Fig. 2: KC identification and control process

 

KC monitoring and control

KCs are measured and controlled using statistical process-controlled techniques to ensure that each KC feature is held as close to its nominal value as possible to reduce or eliminate the cost of quality. An ultimate goal of the enterprise should be to reduce or eliminate KCs. In moving toward the elimination of KCs, continuous process improvement efforts must be undertaken on the factory floor to improve KC performance. As the underlying process capability approaches Cpk >= 1.33, consideration may be given to drop this feature from the active KC list.

Problem statement and hypothesis

At Northrop Grumman Aerospace Systems (NGAS), we saw an opportunity to reduce labor cost by using minimal capital investment to automate the report-generation process. Analysis of time spent on these activities pointed to the fact that the manual report-generation process took just as much time as was spent acquiring the metrology data. The new process automatically uploads the KC part-acquired data and then performs the deviation analysis, using the already embedded tolerance values for the associated part. The CATIA model is imported into the application software, a small section at a time, to mitigate any computer memory issues.

Current process

The current process, described in figure 3, involves the use of laser trackers operated by two mechanics. The scanning of the KC surfaces begins with the laser tracker alignment.

Once the scanned data are acquired in comma separated value (CSV) format, it is converted into data template (DAT) format before importing into the Metrolog software. Nominal values for each KC are manually entered into the system to conduct deviation analysis for each of the 65 characteristics.

Additional file format transformations are required (PDF and TXT) before the file is ready to generate Excel or PowerPoint charts required by the customer. All of these activities are manual and increase both labor costs and product cycle time.


Fig. 3: File transformations: current process

 

Improvement opportunity

The improved process eliminates the need for manual file format transformations (see figure 4). Once scanned measurement data are available, they are imported into the application software along with the respective portion of the aircraft. A script is then written that loads the respective tolerance values for the KC deviation analysis. The resulting data can generate the required customer reports with no file format transformations required.


Fig. 4: Improvement opportunity

 

Project objectives

In 2010, NGAS initiated a system development project for automated data analysis and report generation using commercial-off-the-shelf (COTS) application software. The cost center selected for testing this software was an ideal candidate for initial implementation because it was at the end of the assembly line and prior to the aircraft shipment to the customer. The project goals were:
1. Develop an automated system that can eliminate file format conversions of the scanned data
2. Perform deviation analysis against the model
3. Capture screenshots required to illustrate KC performance
4. Create report generation in both PowerPoint and Excel formats.

System design

The following requirements were identified before development work began.

System requirements:
• Be affordable in terms of initial software development and periodic maintenance cost
• Be user-friendly and easy to learn
• Provide a work-around for those KCs that cannot be measured because a part is missing or is being reworked
• Issue an error message if the file containing certain KC data is missing
• Provide graphically rich reports for easy analysis and corrective action
• Guarantee data integrity from data importation through the report generation process.

Report requirements:
• Flag alarm conditions using visual and audible alarms during the measurement process
• Perform archiving of all pertinent data by each part, version, and serial number of the aircraft
• Create reports using the same formats as the manual process
• Provide capability to print the report or view online
• Provide reports on quality inspection "buy-off" of the features per inspection requirements

 

New automated process

The new report-generation software, developed in less than three months, has Report Plan and Execution as two major components. The Report Plan function is basically a menu of all parts and KC features that are required in the data analysis and report generation. The menu can be revised as changes occur in the KC list. As shown in figure 5, the green area indicates that the KC is within tolerance, yellow indicates it is approaching the limits, and red indicates it is out of tolerance. Those areas that remain white indicate the fact the scanned data are missing for analysis.


Fig. 5: KC performance

 

The scanned files, in CSV format from laser trackers, are directly imported into the application software, eliminating several of the file transformations required in the old process (see figure 6). Each KC datum, i.e., part number with its corresponding surface relationship, is defined. The tolerance values for each KC area are already loaded into the system for deviation analysis, which saves a considerable amount of time.


Fig. 6: Automated process

 

The operator creates the PowerPoint report with a single mouse click. The process opens the NGAS PowerPoint template and uploads screenshots and other graphic-rich data (see figure 7). Screenshots are grayed out to indicate missing data. The floor team then analyzes the results and takes corrective action.


Fig. 7: KC areas

 

Figure 8 provides a graphic of the KC performance summary. Deviation data are presented in balloon and bar formats for ease of reading and understanding. The KC status to the right shows which KCs are in or out of tolerance. The white space shows the missing data.


Fig. 8: KC performance summary

 

Conclusions

1. The previous manual method of report generation was time-consuming, requiring several file transformations. In some cases, time spent for report generation was as lengthy as it was for data acquisition. Substantial savings can result from implementation of an automated report-generation system.
2. Software developed for this project can be used for applications in other cost centers by examining data flows specific to that cost center and by creating measurement maps to guide the mechanics for report creation.
3. The desirable software strategy is to implement a common software data analysis and report-generation platform throughout the assembly line (see figure 9). The results of such a strategy are reduced software development, implementation, and training costs.


Fig. 9: Software strategy

 

4. The automated report generation can be leveraged across the entire production line. For migration, all that is required is to create a measurement plan for each cost center and identify its specific reporting needs, parts used, and associated tolerance requirements.
5. The NGAS KC analysis and automated report-generation system is transparent to the metrology head used for data collection. In the future, if the data acquisition system is changed, there will be practically no impact in the way the data are analyzed and reports are generated. This approach gives the company desired flexibility to change the method of data acquisition as advanced metrology solutions become available.

Acknowledgement

The author wishes to thank the following persons who participated in the development of the automated report analysis and generation solution:
• Ernie Huston, president, Verisurf
• James Edwards, sales manager, Verisurf

Discuss

About The Author

Ken Vakil’s default image

Ken Vakil

Ken Vakil, manufacturing technology engineer at Northrop Grumman Aerospace Systems, has more than 45 years experience in manufacturing engineering, industrial engineering, and financial systems. Some of his current projects at Northrop include development, justification, and integration of advanced metrology systems that are focused on meeting engineering requirements, making the company products affordable while ensuring that newly implemented manufacturing operations are user-friendly and value-added.