As a PI administrator I would like to be able to create multiple instances of "PI Analysis Service" windows service for these reasons:
- avoid new potentially problematic analysis from affecting the ones that are operational and working correctly by setting one production instance and one test instance for example.
- every instance could have its own log files and we could move some analyses to a new instance to troubleshoot them
- make possible to distribute the resources usage for asset analytics on more than one server

As a developer, I would like the analysis service to support custom functions that I write using AF SDK.

Allow for Asset Analytics to support event-weighted summary functions (as opposed to time-weighted), such as average and count. This will allow for easily aggregating changes in digital tags, discrete sensor readings (e.g., bucket count), etc. This would allow Asset Analytics to easily perform many of the event-weighted functions of PI Totalizers.

As a a user of Asset Analytics I want to see the calculation dependencies for an analysis so that I can understand which analyses are dependant on an analysis and view potential circular references

As a user configuring a Rollup analysis, I want to be able to apply a filter expression on the inputs so that the Rollup only includes inputs that satisfy a condition.

Note: This was previously Enhancement 114349.

As a user of event frames, I would like the ability to be able to set a time true condition for the end trigger of an event frame analysis, similar to how this functionality is available for the start trigger

Sometime you want to do an analyse only when one or more conditions are true, if the are not true you want to exit the analyse, so steps below the conditional exit are not evaluated/executed. The conditional in Analytics should look like:
 
if condition = true then exit analyse
 
At the moment this is only possible with the function NoOutput(), but then the mapped attribute should be connected to an PIPoint and you have to fill in a else condition. All rows below the conditional exit will not be executed.

As a user of analyses, I would like to have a Percentile function so that I can filter out atypical events when doing summary calls. The function would look like this: Percentile('attribute', 'timerange', value). This is related to enhancement request WI 44140.

As a quality engineer, I need the ability to calculate Six Sigma SQC statitistics for key quality attirbutes. 

Cp, Cpu, Cpl, Pp, Ppk

Currently in the analysis service configuration there is a parameter called “AutoBackfillingEnabled” which affects the analysis calculations in a global basis. It would be great if it were possible to disable the auto backfilling to individual calculations.

As an analytics admin, I would like to be able to add more complex code such as Loops, Arrays, and other advanced logic to AF Analytics calculations SO THAT I can perform more complex and custom calculations while leveraging the scheduling and other features of AF Analytics.

As a user, I would like to have the option to automatic recalculation of event frame for out-of-order input events in AF analysis. The option is available for expression analysis in PI Analysis  2017 R2 but not for event frame generation analysis.

Currently PI Analysis Service has a built-in Statistics tool in PI System Explorer.

It is able to show statistics regarding Evaluation and Calculation Lag as well as Skipped Evaluations and some other important information.

PI System Explorer could run these statistics periodically and notify the user (in the Management tab or in the Analysis tab itself) whenever a PI Analysis has its parameters out of certain values (could cause a bad performance)

So that the users could avoid performance issues more easily.

I would like to have monthly and yearly periodic trigger for Analysis-Expression and Event Frame Generation. Currently I have to use the daily trigger and check in the analysis if the month or the year changed. In case of “false” there is no output. In the case of “true” there is an output. This works but it is circuitous.

We currently have Event Frames generating for machine faults, with child event frames then tracking things such as maintenance team response time etc.
Leveraging this data in excel reports is proving difficult however due to the Child Event Frames having the same template and naming convention as the parent.
If Child Event Frames could have custom naming or different templates it would be extremely valuable