When mouseover on a output attribute analysis icon the analysis is displayed in the tooltip on a single line. The analysis will quickly become unreadable.

It would be nice if the analysis could be presented in a readable way.

Ability to allow extensions in PI AF Analysis template, in order to modify specific element analysis.

I would like to be able to reference the parent of a element reference. As it currently stands, it is only possible to reference the original parent element of an element.

As a user, I'd like to have a Calculated Data option when i select the Time Series Data of an attribute. This would let me pick out things like 5 minute averages, 1 week range etc etc.
It would put data in end users hands without having to request it from the engineering team or learning PI Datalink.
Right now, its only Sampled Data or Archive (Raw) Data available.

In many UseCases we need more programability functions in PI. As example loop functions. We like the AF Framework, beacuse if you search in case of a problem you are very fast. But the functionaclity is unfortunatly not very rich.

Please integrate some more functions or integrate NodeJS, Phyton or
other scripting capabilities

Transfer Saved searches: This option will allow transferring event frames "saved searches" from one computer to another. For customers who have many search criteria, it is time-consuming to rebuild again on another PC.

As a PI AF administrator or power user, I would like to be allowed to continue loading content from a server library into a database, even if a UOM conflict is detected.

Upon finding a UOM conflict, the library import feature does not allow me to ignore the error and continue loading the content (screenshot1.png), in contrast with the file import feature (screenshot2.png).

Tech Support case #00957957 concluded that this was not a bug.

As a PI Admin in a highly regulated environment, I need to demonstrate that the AF Servers in my production and qualification environments are identical.

OSIsoft does not recommend comparing PIFD databases directly, and there are small differences in XML exports (perhaps influenced by which information has been loaded recently).

I would appreciate a tool that I can use to programmatically demonstrate that two AF Servers are functionally identical (that is, they'd provide the same client behavior and the same outputs given the same inputs).

Currently, when creating an Attribute in PI AF using "String Builder" or "None" Data References, the system does not update the timestamp associated with the attribute's value upon value change. When the end-user looks at the timestamp they see a static value: "1/1/1970 00:00:00" which is not representing the last time when the value has changed and thus very confusing. I believe that every value change should have a timestamp associated with it regardless of the Attribute type.

If you try to do Value Retrieval Method (like Min) of a Digital type tag, you will get an error saying non-numeric data. AF should be smart enough to realize that the data is actually a set of integers and not try to read the string representation.

Currently when viewing Time Series Data for a numeric attribute value type (e.g Double, Single) in PI System Explorer, sorting the value column will sort the results alphabetically (e.g. 1, 10, 2, 20, 3...) and not by size (e.g. 1, 2, 3, 10, 20). PI System Explorer should be able to sort Time Series Data for a numeric attribute by the size of the value in addition to or instead of the alphabetical order of the values as strings.

The assignment of a PI tag on which the currently logged in user has no authorization as PI point reference to an AF attribute must not (as currently) be possible.
It is possible to bypass the PI Archive authorization mechanism (point and data security) by using such attributes in AF-Analytics.

The error messages that appear when saving in the System Explorer should not be allowed to be skipped.

When moving an AF element within the hierarchy in PSE using the "Shift key + Drag and Drop" approach, it can cause to create an Element reference if the release of the "Shift key" is done before dropping the element.
Having the option of right-clicking on the element and select "Move" would eliminate issues with pressing/releasing shift key and dragging/dropping the element and creating Element References by mistake.

Looking for a way to have multiple reason types for AF attributes.

Currently, AF search results can only be sorted by a single field. It would be useful to have the capability to specify 2 (or more) SortField values so that the order of results can be guaranteed in the result of a tie.

For example, being able to sort Event Frame Search Results by both StartTime and EndTime.

large value display as 1.23 E+008 format instead of the orginal value in PSE Pop-up Trends

if the trend can follow display digits, it will not show scientific nonations