All entries for January 2011
January 20, 2011
After the lecture about different reliability tests I had one idea based on reliability test of a process.
The example was that if we have one machine if we note down its every breakdown we can calculate Mean Time Between Failure which can help us plan maintenance activities.
Now in my idea assumes taking every out-of-control state of a process from the SPC study as a failure (instead of taking breakdowns) to calculate MTBF. If it's possible to do it this way (!), we could calculate MTBF for the process and then carry our Wiebull analysis. If the test's assumption was fulfilled we could calculate 'beta' and we would get to know the life stage of our process! In my opinion this would be really valuable information, as i.e. if we leanred that the process is in the 'wear-out' stage we could plan an earlier maintenance or consider buying a new machine etc.
The question is - would it work this way? :)
January 17, 2011
During the lectures we talked about risk management and risk prioritization. We were told that companies normally use scoring system of 1,3,5 or 1,3,9 for each category and then multiply them to get a list of the most urgent issues.
One of our colleagues said that in his company they use scoring system from 1 to 10. Then a lecturer said it might be sometimes confusing as people may have problems if they should give a score of i.e. 2 or 3. Thus 1,3,5 scoring is easier (weak, average, strong).
In my opinion the 1-10 scoring makes sense as well though, cause if we use the first type and if we had just 2 categories we may end up with scores of only 25,25,25,25,25.... no prioritization benefit! And if we had more numbers to choose from we will have better range of scores (but then even slight change from 1 to 2 may cause that the issue will fall 10 positions up or down in the priority list).
And finally I said that in my company the scoring system (maybe not exactly only for risk studies) depends on more categories which include costs of parts, warranty claims and else (which I can't remember now and the rest is a secret ;) and in this way everything is sorted. Hearing that there were some comments that cost category is unnecessary as if this was a customer safety issue we have to deal with it no matter the cost.
I agree with that but then I would say that others missed the point of my comment. We don't use it to see how much it costs to deal with the issue BUT having a priority based on cost (imposed on the company) we know where to allocate resources to get most savings. At the end of the day, other companies would carry out cost analysis as well...
January 06, 2011
After half a day of lectures and a day of research I started becoming familiar with DFSS.
At first I had an issue with DFX and it's place in DMADV methodology - I had no idea when it can be used. But after reviewing notes I found out that we use DFX in the Analysis stage. But this was irrelevant.
What was really interesting was the originality of DFSS. During my Bachelor's degree I studied Pugh's Total Design Methodology. It was focused on an efficient product development starting from customers' expectations collection, developing conceptual designs, selecting and creating the best one, manufacture stage and market introduction.
Now that's all what DFSS consideres as well (as far as I'm aware at the moment). I remember that in the Total Design concept we were encouraged to use QFD, FMEA, TRIZ, DoE and else.
As far as DFX is viewed, Pugh's concept assumes creatin a Product Design Specification which includes 34 areas that have to be considered to ensure the design is defined appropriately. The document has similar areas as DFX (however I'm not sure if with PDS we can make the same level of analysis as in DFX would assume. I'm not sure as I didn't use PDS to this extent and I can't remember anything about that from the theory now)
Anyway the Total Design is very similar to DFSS. I was wondering what was first. Pugh published his concept in 1991 but I can find when DFSS was introduced...
Figures are from Wikipedia, I'm not gonna reference it as I should be working on the presetation anyway!