QA6640 Chat Room Transcript January 31, 2007

 

Prof. Atkins 20:18:18
The chat room is closed for this evening. It will reopen next Wednesday at 7:35 PM.


J Scott 20:14:50
Good Night All


Rich Weaver 20:12:23
good night!


Prof. Atkins 20:11:48
If there are no questions, we can close the chat room for this evening.


Prof. Atkins 20:05:49
The thirty-minutes are now up. Responses posted after this message appears will not be evaluated for chat room participation credit.

If you have a question, you may ask it at this time.


el hamraoui hanane 20:03:34
Hanane El Hamraoui
helhamra@spsu.edu
I used to manage a clothing store where we get lots of cloths in a daily basis from national clothing manufacturers (suppliers)
In our warehouse a quality control process has to be done to detect any nonconformities in the lots received. For this purpose the quality department use the p charts to estimate the percent defective that does not satisfy one or more of the product specifications of all lots shipped.
We use the percent nonconforming as 100% times the fraction nonconforming of all lots shipped to the warehouse.


komic 20:02:57
The company I work for is Cardinal Health Snowden Pencer Products. We manufacture surgical instruments. Some of the products is packaged in the Clean Room environment. The Clean Room is classified as ISO Class 8. This mean that maximum permitted airborne particle (particle size 0.5 micrometers) per cube meter is 3, 520,000.00.

After initial classification, the air cleanliness must be controlled and maintained. To do so we established the upper control limit by collecting the data over 6 months period.
Samples were taken once a month at nine locations within clean Room.

We used x-bar chart for this purpose. As per Bossert text “X-bar and R-charts are used when there are subgroups or data that are collected”.

We took samples at nine locations as determined for the size room we had (aprox 52 sq.meters).

At each sampling location 5 single volumes were taken.
The particle concentration at each location obtained from the measurements were recorded.
Average particle per cubic ft was calculated for each location.
Overall mean of the averages were calculated.
Average of averages was calculated after six months data collection.
Standard deviation was calculated for the six months averages.
Upper control limit was determined as average of averages + 3 st. deviation.

The benefit of using this chart is simple control of the air cleanliness of the Clean Room. Each month we are taking samples at all nine locations, calculate average and monitor cleanliness by using this chart and established control limit.


Michael Ginn 20:02:38
Michael Ginn
Email: ginnmichael@hotmail.com


1. X-bar and range/standard deviation charts

The company I will reference to is a tier one automotive supplier (Transmission Supplier to the Big-Three). I worked there as a SPC technician. This company is a fairly large plant (200-250 employees). They supplied a total of 12 different models of transmissions. Being a SPC technician I directly controlled and monitored the control charts used throughout the plant.
Through out the plant we used SPC to control all critical characteristics. These characteristics could be a bore position, bore diameter, or even a depth/length. These X-bar and range/standard deviation charts we used so any out of control feature could be caught faster. The chart has a target (x-bar) and upper and lower control limits. The target and control limits would only be adjusted if a process change justified a recalculation. Very rarely was a chart recalculated. The operators of the processes where trained that process should stay with-in its limits. Any out of control processes were shut-down for a pending investigation. Also, trends or runs would justify a process being shut-down. A trend or run is where an operator would seven or more points moving in the same direction. The methodology for this is that if the process continues to run in the same fashion the trend/run will eventually go out of control. Upon implementation of the X-bar and range/standard deviation charts processes where controlled much better. This impart lowered scrap costs also. The charts made the operators visual aware of how their machine was running. Before the charts the operator could not tell you how their machine was running other then it’s in tolerance. A good example of one of the features controlled by X-bar and range/standard deviation chart is the dowel bore. The specification for this bore is 12.037 +/- 0.008. The x-bar would normally be around nominal (12.037) because the bore was cut with a diamond reamer, and the limits would be +/- 0.006. If I remember correctly?


Joel 20:02:23
QA6640
Joel Centeno
jcenteno@twt.com

Company: Boston Scientific
Boston Scientific engages in the development, production, and marketing of a variety of medical devices for cardiovascular, endosurgery, and neurological use. The specific division I worked for was the Grinding & Straightening Subassembly Guidewire Business Unit with focus on endoscopy, urology and peripheral vascular medical device product lines.

Type of Control Chart: p-chart
This type of control chart looks at the percent defective of all lots shipped

Example: Nitinol Guidewire spool monitoring

In the unit that I worked, we manufactured about 6,000 guidewires per day. Before wires were ground, they needed to go through a straightening process and heat treatment to achieve the desired mechanical properties. The straightening operation began with a spool of wire (about 100 yards of wire wound in a single spool). With such mass production, spools of wire came in from supplier at a very fast rate (almost 500 spools per day per supplier). The receiving inspection department sampled each lot of units shipped using a specific sampling plan. I believed they sampled about 50 spools or so. This was not a destructive test and it was relatively simple acceptance criteria. If they starting unwinding the wire from the spool and upon the first few yards of wire they saw kinks, bends, or inability of the wire to retain memory the spool was rejected. Based on the amount rejected and the sampling plan, the lot of 500 spools would pass, be inspected 100%, or returned to the vendor. On a weekly basis, the quality engineers would gather the data of the of defects found in one lot and would accumulate the data for a 3 month period for all lots for one supplier and created p-charts for each supplier (about 3 I believe). The data serve greatly in determining the consistency of each supplier and making decisions of any corrective actions needed for supplier to send us better material. It also allowed us to allocate resources in working with the suppliers who showed the most problems.


Natasha_Romero 20:02:11
Natasha Romero
nromero@spsu.edu


p Chart

1. I currently work for Sandia National Laboratory. We are a contractor to the Department of Energy and Department of Defense. Our mission is national security.
2. The p chart as described by Bossert, “is used to estimate the percent defective of all lots shipped.” The use of this enables the customer or in the case of the Shipping and Receiving Department which is discussed below to make decisions about how the process of production or performance. This type of chart can help identify areas for improvement.
3. In the Shipping and Receiving Department measurement and analysis is conducted through the use of an electronic system which collects data about when an item is received at Sandia National Laboratories and when the item has reached its final destination. These metrics are used to produce a p Chart which shows the percentage of time it takes for an item to be delivered for a particular route. This p Chart enables management/supervisors to track and monitor the disbursement of items to their final destination. Through the review of the p Chart management/supervisors could potentially identify a delivery route that is taking longer than average to deliver items to their final destination or a route that is taking less than average to deliver items to their final destination. Management/supervisors can then investigate into the reason as to why these types of behavior are happening and improve their route system so that the delivery time of an item from the time it is received at Sandia National Laboratories and the time it reaches its final destination is as close to the average time deliver time. It could also be discovered that the route for delivery is going to have a lower or high delivery time due to some type of extenuating circumstances such as the route final destination is located at a 20 min drive from the shipping/receiving warehouse in comparison to the other routes which only have a 5 – 10 min drive.

Bossert, James L. The Supplier Management Handbook Sixth Edition. ASQ Customer-Supplier Division.


gstevens 20:01:50
Module 3 Discussion
Glenn Stevens
Email: Anandale@comcast.net
Date: Wednesday, January 31, 2007


Company Selected: Home Depot, this company is a major retailer of home improvement products. Most of their decisions are based on sales data information. That is information is collected daily from the 1800 stores, processes and then redistributed to the business end user community to make decisions.

Chart Type: The chart type selected is the X-bar and range/standard deviation chart. This type of chart allowed data to be collected in lots and averages determined, upper and lower limits (range) and standard deviations derived and the actual or observed data values compared against these threshold values to determine variances. Such variances allow for analysis to determine root cause.

Example: This example examines the use of the chart type in collecting daily sales data for the company’s stores and performing quality measurements to ascertain on time arrival of the sales data and the accuracy of the sales data.

A software program was developed that calculated the moving daily average of the sales for stores based on the sales data for the past 5 weeks for the same day as well as the daily range. That is the average was calculated based on the last 5 same day of the week. Also the standard deviations were also calculated. The sales amount from each of the stores received was then compared against the average for the last 5 days (X-bar). Additionally the actual sales amount for each store was calculated and this had to be within 4 standard deviations. The range was also calculated and compared with the reference range.

Benefits
Because the measurements were automated, stores were easily identified that exceeded any of the three threshold values. Based on the magnitude of the difference between the actuals and the average or the falling outside the defined range, specific actions would be taken. The greatest value was that the data processing department was able to advise their end user customer that the data to be supplied had questionable data before they actually encountered it. These stores whose data were outside the thresholds were contacted to determine the cause. Over time it was seen that certain stores were showing a pattern. For instance it was discovered one store in Hawaii and an old modem which had problems in transmitting the data. In other cases, the variance could be explained. A store was closed because the president was visiting the area, and hence had to be closed impacting sales. Other discoveries revealed a shooting occurred at the store, a fire was at the store, or a weather related event occurred. One problem however was that because the average was moving, the abhorrent values also influenced the new averages. Most importantly, the method allowed for early detection of a problem before the data was passed to the end user customer without knowing, allowed the supplier (store) to be contacted to determine root cause and to take measures where necessary to prevent the recurrence of the cause of the problem where possible.


J Scott 20:00:22
I previously worked for Delta Air Lines in the continuous improvement group. We were examining a bearing machining process due to the fact that landing gear bearings received from the machine shop were either getting stuck during installation or could not achieve the tolerance for the .001 interference fit. We started our analysis by checking each bearing coming off the large lathe that was being used to machine the mating surface. We used the Individual and moving range chart on a initial sample of 12 bearings. The IMr chart chart uses individual points plotted on a chart. Upper and lower spec limits are used to set the limits. The range is measured to determine the variation between points. The individual and moving rang chart showed that 4 of the 12 bearings had exceed the upper control limits at various stages. To bring the process into control, we developed a standardized procedure for machining and checking the bearings. We also started checking the bearings after they had been cooled to about 65 degrees, due to the cooler environment where they were installed. Another run of 12 bearings showed that we had stabilized the process but we had too much variation in the process. We developed a procedure for chucking the bearings in the fixture and also ensured that our gauge was calibrated. After another run of 12 bearings, we had brought the process into a sufficient amount of control. All Individual points were within control limits and the range (variation) between points was also reduced.

The control chart enabled us to more effectively analyze our process without a shotgun approach to problem solving. By using the control limits, we were able to tell when we had sufficiently brought the process into stability. After, the machining process was stable, we were able to make adjustments to the process to improve the machining of the bearings. The results were very favorable with very few complaints. The benefits of the analysis showed that temperature and the lathe bearing fixture was inducing variation into the process.


Rich Weaver 19:59:40
Having the supplier provide data can be beneficial for both parties; the supplier provides documented evidence that their process is in control, and the customer develops a level of confidence that the source is monitoring themselves, and is capable of detecting a mean shift or increased variability, and will take corrective action before discrepant parts are shipped.

Recently, we experienced a quality problem with the front axle drive shafts (FADS). The FADS connect the transmission and the front hubs. The correct diameter of the lip of the drive shaft (where it mates to the transmission) is a critical characteristic. Too large, and the FADS will not properly snap into the transmission. Too small, and there is the potential for leaks or premature wear.

There were a number of complaints in a short period of time about the drive shafts coming out of the transmission. In all cases, the lip of the drive shaft had not been fully seated. There are a number of variables in our own process that can affect this operation; correct alignment of the FADS relative to the transmission is critical. This is complicated by the fact that there are over 30 “body trucks”, and variation in the body trucks can cause improper alignment.

As is often the case, the consensus among the plant floor people was that the problem was with the shafts. (There had been no revisions to the trucks in several months.) Source was contacted, and sent in a representative. The shafts that he examined were within specs.

Things were at a standoff until our plant asked for data from the supplier on the dimension in question. They have a strong SPC orientation, and were able to provide historical data, in the form of X-bar and R charts, which demonstrated exceptional control. The source representative was able to show that the sample means were consistently within the control limits, and that the sample ranges were in control as well. A (GM Central Office) Supplier Quality rep visited the supplier facility, and was able to vouch for the controls in place.

In conclusion, it’s hard to argue with data. They were able to provide proof that there was only common cause variation in the size of the drive shaft diameter over time. Besides providing evidence for this particular problem, their use of data and SPC techniques to monitor their processes was impressive and instructive to the production organization in our own plant.

By the way, the problem turned out to be lack of maintenance, on our part, on the body trucks. The detail that supports the other end of the drive shaft had deteriorated on several of the body trucks, allowing the FADS to align at an incorrect angle.


Gleiter 19:59:19
Kimberly Gleiterkgleiter@hedonline.comSC Packaging produces foam molded parts. One of their critical specifications is the weight of the part. The weight can tell them if the correct density of material has been used in production. They utilize X-bar and range/standard deviation charts to graph the results of their subgroups. So for example, they would weigh 5 parts and graph the average of the results. The downfall to this process was that it was not graphed in real time. The operators did not have access to a computer, so they would record the information manually and later someone would graph it electronically. By the time the data was entered and the out of control occurrence was detected, nearly an entire shift of product may have been run. Once it was graphed however, they could quickly detect whether something out the ordinary was going on (typically the press was hooked up to the wrong bag of material). Because it was such a rare occurrence that the wrong bag was used, management did not see the benefit in tracking this in real time (providing computers to production personnel), but they continued to record the data in order to meet a customer requirement.


Brad McGuire 19:58:45
QA6640 - Brad McGuire
bmcguire@spsu.edu
1) I will be using my organization – Lockheed Martin Aeronautics Company in Marietta, GA
2) I will discuss the u-chart (total defects per unit) as it relates to the communication process between QA and C/A Engineering in F-22 Final Assembly.
3) As part of my regular job assignments, I’m tasked with performing a Radar Cross Section Walk-Around in final assembly to ensure all aircraft panels are installed with acceptable mismatch and gap per specs and engineering drawings. Once I have completed this walk-around process, I ask QA Inspection to document all of the out-out-tolerance findings (if any) and submit these rejections to MRB Engineering for repair dispositions. I then, submit the following to Corrective Action Engineering for feedback purposes: 1) rejection document number, area of aircraft – panel numbers, out-of-tolerance measurements, and the history of this rejection (if any). In addition to this, I also submit the total number of nonconformities (step or gap) on the aircraft to QE and C/A Management via u-chart that illustrates the number of findings per jet going back approximately 50 ships. This u-chart data is extremely useful in that I’m able to easily extrapolate a control chart with upper and lower limits. This chart identifies whether or not the RCS process is in or out of a state of statistical control and dictates whether or not Corrective Action has been validated. Other benefits include: simplicity of maintenance, shows logical sequence of jet rejections, trends can be easily recognized, and is the primary tool used when reporting overall percentage of rejections, cost and recurring problem areas. And perhaps most of all, as I work with Production and Corrective Action Engineering toward reducing these rejections, management can easily recognize this by seeing the u-chart defect numbers decrease.


Chris Wellman 19:58:27
Christopher J. Wellman


Chart Type: X-bar and range/standard deviation charts
Company: Kollmorgen Corporation

Kollmorgen is a Contractor to the Department of Defense; we design and build complicated Sensor Systems as large projects in a very low quantity. From initial contract kick-off to shipment of a system it typically takes about 1.5 years. Generally, we work on no more than a couple of systems each year for a given program. The business is divided into three business areas.

We use the X-bar and range/standard deviation charts, which you can use to identify to “unusual occurrences that cause an out-of –control condition” as discussed in Pages 34 to 35 in the Bossert textbook for a given business area.

This chart quickly shows if our process for handling problem, that is our ability to quickly resolve them. As you can imagine with Thousands of parts that make-up a system it is important to quickly close-out problems or something that was not critical can quickly end up on our project critical path. So, the visibility provided by our -bar and range/standard deviation charts for problem management is important to identify when we are not keeping portions of the project moving along on schedule.

The visibility of the charts allows us to prevent schedule problems which might otherwise cripple a project. The three business area representative meets regularly to share common problems and solutions. This way we can bring additional resources from other groups to bear on areas that are not in control as demonstrated by the charts.

 


Alan Dial 19:52:51
Alan Dial
Dalzell22a@yahoo.com
I have chosen to discuss GE Contractual Services (GECS). GE Energy is a primary business unit of General Electric. GE Contractual Services is a division of GE Energy. GE Energy provides power technology solutions for Independent Power Producers (IPP) and public utilities. The business provides consulting, sales of turbine packages, installation of such packages, operation and maintenance, as well as service contracts for the warranty of these packages. Contractual Services provides long-term contractual and financing solutions for companies on purchased GE OEM equipment.
I chose to discuss how our business uses the I&MR Chart for chemical analysis. These charts are excellent for tracking long-term variation changes. Because they use a single measurement for each data point, they are not a tool of choice where measurement variation is involved, such as with part dimensions. They work well with temperatures, pressures, concentration, etc.
The I&MR is used at our Operations & Maintenance facilities. In this portfolio, GECS operates and maintains the power plant for a contractual customer. GECS performs a chemistry analysis on the Boilers at the combined cycle facilities. GECS chemistry technicians will plot the data in an I&MR chart to monitor the change in pH and Phosphate concentration. Because there is little variation in the measurement due to utilizing highly calibrated and automatic equipment, I&MR is a good tool for measurement. As well, an individual point is being measured, not a subgroup.

 


Prof. Atkins 19:34:31
CHAT ROOM DISCUSSION QUESTION:

You have thirty-minutes to answer the following chat room question:

On Pages 34 to 35 in the Bossert textbook there is a very brief discussion of the six basic types of control charts:

1. X-bar and range/standard deviation charts.
2. Individual with a moving range charts.
3. p charts (percent defective charts).
4. c charts (defect charts).
5. np charts (number of rejects charts).
6. u charts (defects per unit charts).

Select any ONE of the above six types of control charts and discuss it in detail as it applies to a SPECIFIC example that you are familiar with.

You may discuss your example in reference to any ONE of the following:

The company you now work for, or
a company you have worked for before, or
any company you are familiar with.

Your discussion should cover the following three issues:

1. Briefly describe the company you have chosen.
2. Briefly describe the type of control chart you will discuss.
3. Use a specific example to help explain the BENEFITS derived from the control chart, and include as much detail as possible without exceeding the thirty-minute time limit for your response.

It is suggested that you type your answer into a Microsoft WORD file. When you have finished answering the chat room question, then you should copy and paste your entire answer into the Message window above. Then click the [POST] button. Your answer will then appear in the chat room for everyone to read.

You MUST post your answer in the chat room no later than thirty-minutes AFTER this question first appears in the chat room.