1. Problem solving tools 1 Problem solving tools We need tools to: –improve designs: increase the chance that we are making what the customer wants. –identify problems: identifying the problem(as well as prioritizing problems) is half the battle. –prevent problems: key to quality systems is prevention. –Note on these tools and general simplicity Who needs to use them?

2. Problem solving tools 2 Linkages Easiest one of the semester: These tools can be used to make any process better - and in the case of SPC can be used to prevent problems from occurring in any repetitive process.

3. Problem solving tools 3 Variance exercise Hold the slide (envelope) a foot above the target. Slide a coin down the envelope and try and get it to stop on the center of the target. –In other words your process is sliding a coin down a slide onto a target. Do this 5 times to set up the process –once set-up repeat 10 times and record results –then do 10 more runs with another coin recording the results Change the person holding the slide - and do 5 set-up runs again –do 10 runs and record the results As a group write down all the sources of variance in your results

4. Problem solving tools 4 Variance Variance is one of the major problems in a process (and it is the cause of all quality problems) As we reduce variance things happen as expected- so if we have no variance in a process it performs as expected (designed) in other words if we have a design that customers want and then we execute the design with no variance we make $$$$ So our tools are used to identify and remove variance (after we design a good product)

5. Problem solving tools 5 Improving designs: Taguchi Taguchi aims to design processes that produce products that are right on target. Why target values over specifications ? –Tolerance stack-up Initial quality surveys verses long term quality surveys - especially Toyota –Specifications allow variance - a target is an attempt to have no variance - which means what ? There are costs for being above or below the target value

6. Problem solving tools 6 Taguchi continued As we get further from the target value the loss increases: –tolerance stack-up –increased opportunity for scrap or rework –lower performance –wasted material –etc. Goal is to continuously improve process with an aim of one day producing everything on target - no loss

7. Problem solving tools 7 Identifying problems: process charts Problems verses symptoms ! If our processes are not under control (to be defined later) we can not produce parts that are on target every time because we will have to much variance. In order to get a process under control and then improve it we need to know what the process is.

8. Problem solving tools 8 Process charts 1) Show us what we are doing ! –identify data collection points –isolate problems –identify areas for improvement 2) Once we understand what we are doing and where the problems are we can design a new process with (hopefully) less variance and or waste.

9. Problem solving tools 9 Process charts - an example

10. Problem solving tools 10 Process charts - class example

11. Problem solving tools 11 Questions for process chart Where are there inefficiencies in your process ? Where are quality problems most likely to occur ? Can you change the process to decrease these problems ? To reduce waste ? –Are we ignoring a class of waste in these charts?

12. Problem solving tools 12 Cause and effect diagrams A tool used to identify all possible causes of a problem. Forces people to look for alternative explanations for problems. Also called fishbone or Ishikawa. First step in problem identification - brainstorming.

13. Problem solving tools 13 A C&E example missing bottlecaps Material defective late Machinery adjustment broken capper wrong program Manpower absent lack of training substance abuse Methods scheduling machine loading

14. Problem solving tools 14 Class C&E MaterialMachinery Manpower Methods

15. Problem solving tools 15 Pareto charts - histograms A method of prioritizing problems. 80 / 20 rule We want to eliminate all problems- but we should deal with those that occur most often and or have the highest costs (losses) first. –What should determine the order of problem elimination?

16. Problem solving tools 16 Pareto Example late caps 45 % absent operator 25 % scheduling 15 % wrong program 5 % all others 10 %

17. Problem solving tools 17 Inspection Can we inspect quality into a product ? How do most companies do inspection –QC function at end of “line” –Timing of feedback? –Blame games? Three key problems with inspection –Happens after the fact - so it can not prevent problems - and inspections use for preventing future problems is limited due to time lag and focus on outcomes Harley Davidson –It is a process and hence can be done wrong Libbey Glass –Focuses on the product / outcome not the process

18. Problem solving tools 18 SPC - A way to inspect to prevent problems Statistical process control - focus on process not outcome. Done in real time by operator –why does this matter? Costs less than tradition inspection methods Can be used to prevent problems

19. Problem solving tools 19 100% inspection or SPC 100 % –high costs of failures –new product, process, tools, set-ups, etc. –non-repetitive tasks SPC –repetitive tasks –low costs of failure

20. Problem solving tools 20 Statistical Process Control (SPC) A process is in control when the only sources of variance are common (natural) causes. A process out of control exhibits assignable variance (sometimes called special cause variance) We determine control by using sampling and statistical methods. Description of SPC process

21. Problem solving tools 21 A sample control chart UCL LCL Target

22. Problem solving tools 22 A sample control chart (2) UCL LCL Target

23. Problem solving tools 23 A sample control chart (3) UCL LCL Target

24. Problem solving tools 24 A sample control chart (4) UCL LCL Target

25. Problem solving tools 25 A sample control chart (5) UCL LCL Target

26. Problem solving tools 26 Basic decision rules The concept of control is that statistically we have only random variance. So any point outside the control limits is out of control Any pattern is also an indication that we have special cause variance. –Patterns are usually considered 4 points in a row that can be all on the same side of the center line all part of a trend There are other patterns as well (the shift change pattern)

27. Problem solving tools 27 Key points for SPC Problem prevention tools - why? Statistically derived Over time we will tighten the control limits as we reduce variance in the system - does this mean someday all points will be on the center line?

28. Problem solving tools 28 Does control = good quality? We have been assuming that processes are actually capable of meeting conformance needs. This is not always true –i.e. you need to be accurate to the 1000th of an inch (pretty common tolerance) and you are measuring with a tool with.1 inch increments –If our process is not capable of meeting conformance needs it can be in control and still producing poor quality To be sure that we are meeting conformance needs a process has to be capable and in control

29. Problem solving tools 29 Tools conclusions We have many tools to improve any process If we have no variance we execute what we designed and make $$$ –if we have a good design SPC is a problem (variance) prevention tool as opposed to inspection which is a variance identification tool (and a bad one at that) I will mention this again but we offer a class in managing quality that covers these topics in depth