We are aware of the interest in this functionality but at this time it is not in our short-term plans to implement.
Ryan Mcerlean (OSISoft) ,
Why the hold up? OSISoft has been promising this for a long time and the community is clearly asking for it being that it is at the top of the list. We simply need the ability to place UCL, LCL and the ability to place USL and LSL along with Cp, CpK, PpK, sigma capability, etc...onto our process graphs for filtered data such as when our plants are online vs offline.
Currently we are only carrying Processbook because of its SQC capability. If this was in Vision, we would've mass adopted it by now. Is the concern that OSISoft would lose Processbook license revenues as folks would drop it to mass adopt PI Vision? or is it that something else (REAL SOON) is coming? Such as the ability to do these calculations in AF that can be places into an existing element template somehow?
Good to hear. Here is something I HIGHLY recommend for SQC in PI Vision:
I have an element template for all my assets. All of the attributes in my element template reference PI tags. Those PI tags record data when the plant is on and off. When I run an SQC analysis it conducts SQC on all the data including downtime data. Therefore the analysis doesn't work.
The work around is that in the SQC there is a filter expression one can type in to only look at uptime plant data. The problem is that I cannot utilize the "Plant Run Status" attribute inside of my element. I have to use the individual loop name for the asset. This removes the ability to move through my assets and having the filter expression reference the element attribute of the run status because it is always looking at a pi tag. The pi tag filter expression is static and is now applied for a different asset on my new asset. This doesn't work.
Therefore, I can't use the SQC and filter my data for uptime only without removing the ability to access each of my assets using the same visual layout.
I recommend adding element attribute filter expressions in PI vision so that one can clean up the data as its being fed into SQC.
In fact, it would be nice to have an element template become a filter in itself in that the data portrayed out of the element template is essentially a filtered version of the raw data that is input into the element template. This would allow me to already have filtered data coming out of the element for my asset and would not require the SQC to filter it out.
I hope this makes sense because its needed.
As an end user and manager of PI over all of our sites facilities, SQC/SPC are the way we will be controlling all of our processes. We will not adopt PI Vision across our operations without this feature because we depend on these charts.
If PI decides to develop SQC into PI Vision, one thing to note is that processbook currently calculates CpK values. However, the CpK value is actually the Cp value when a centerline value is typed into the SQC control parameters.
If only a USL and LSL are typed in, without a centerline value, then SQC calculates CpK.
The problem is that the program says CpK no matter what although its Cp when the centerline value is typed in.
This is confusing.
Lastly similar to how there are videos on how PI stores data into the archives, a step by step example of how PI cleans the data with the sigma data filter to calculate sigma and CpK or Cp would be beneficial. I currently cannot get my own statistical calculations to match PI when data filter is turned on because PI filters the data in a way I do not understand and cannot find any documentation on it.