Attribute Gage R&R Basis for Applications - Personal Perception Criteria

[favqty]

Hey everyone,
I am writing this post just because I am feeling that Att-R&R is not a thing that works in my process. We make car's carpets and sound barriers. All my visual characteristics could be controlled just compairing the sample against a pattern (Colour, texture, other attributes). I do not understand how can I align (statistically) the criteria of something that is pretty much driven by personal perception.

I do understand that it is very usefull and indeed a need for gauges but I can not find the reason to develop such a methodology to align operators perception.

I would like to some of you share your thoughts about this MSA requierement and also gives some product characteristics that works as an examples for applicability.

Thanks in advance to everyone.

 

[jrubio]

Hey everyone,
I am writing this post just because I am feeling that Att-R&R is not a thing that works in my process. We make car's carpets and sound barriers. All my visual characteristics could be controlled just compairing the sample against a pattern (Colour, texture, other attributes). I do not understand how can I align (statistically) the criteria of something that is pretty much driven by personal perception.

I do understand that it is very usefull and indeed a need for gauges but I can not find the reason to develop such a methodology to align operators perception.

I would like to some of you share your thoughts about this MSA requierement and also gives some product characteristics that works as an examples for applicability.

Thanks in advance to everyone.

I think AITF believes the same as you and wrote:

All the initial Studies Capabilities are focused on variables none attribute.

Try to pass your attribute var to none attribute variables and establish a measuring method

whether not:

For attribute variables there are technical method such as:

Artificial Intelligence and neuronal nets integrating in special cameras to do this job instead of operators, before a previously training or pattern are entered.

Links for neuronal nets and pattern recognition:

http://elsmar.com/Forums/showthread.php?t=17556

Cheers.

 

[favqty]

I think AITF believes the same as you and wrote:

All the initial Studies Capabilities are focused on variables none attribute.

Try to pass your attribute var to none attribute variables and establish a measuring method

whether not:

For attribute variables there are technical method such as:

Artificial Intelligence and neuronal nets integrating in special cameras to do this job instead of operators, before a previously training or pattern are entered.

Links for neuronal nets and pattern recognition:

http://elsmar.com/Forums/showthread.php?t=17556

Cheers.

Javier,
Thanks for your sooner response having in mind that the post was set during weekend.

Artificial Intelligence is used mostly on high speed processes to help operators (Tell me if I am wrong). In this particular process every step is made by hand. There is nothing that we make at high speed. The operators move everyti¡hing by hand and the heating and pressing is the only thing that we make with machines. So they have the opportunity to check and control the product in almost every single minute. That is why I think we not need to run Att-R&R on this process.

I really appreciate if you can send me the link when you got that phrase from IATF. Thanks in advance for your comments.

FAV

 

[jrubio]

Javier,
Thanks for your sooner response having in mind that the post was set during weekend.

Artificial Intelligence is used mostly on high speed processes to help operators (Tell me if I am wrong). In this particular process every step is made by hand. There is nothing that we make at high speed. The operators move everyti¡hing by hand and the heating and pressing is the only thing that we make with machines. So they have the opportunity to check and control the product in almost every single minute. That is why I think we not need to run Att-R&R on this process.

I really appreciate if you can send me the link when you got that phrase from IATF. Thanks in advance for your comments.

FAV

I really appreciate if you can send me the link when you got that phrase from IATF. Thanks in advance for your comments.

PPAP third edition page 6.
Artificial Intelligence is used mostly on high speed processes to help operators (Tell me if I am wrong). In this particular process every step is made by hand. There is nothing that we make at high speed. The operators move everyti¡hing by hand and the heating and pressing is the only thing that we make with machines. So they have the opportunity to check and control the product in almost every single minute. That is why I think we not need to run Att-R&R on this process.

Bear in mind

Pattern recognition often replaces manual inspections and can be incorporated in mechanical vision systems for industrial inspection tasks. The advantages of pattern recognition are that it is rapid, the same feature is measured objectively time after time, and it never gets tired

If your operator are tired with doing the same checks, Pattern recognition is more reliable, it depends what are you checking and the money you want to invest.