1. OSIsoft Product Roadmap The Server
Chris Coen
Product Manager

2. Functional Groups of The PI System
Performance Improvement
Operational Visibility
Knowledge Management
Product Quality
Lean Manufacturing
Asset Management
Situational Awareness
The
Deliver Information for Meaningful, Better Business Decisions
Visuals
The PI System
The
Turn Valuable Real-Time Data Into Actionable Information
Analytics
The PI System is categorized by three major product functions. They are classified in a somewhat self-explanatory way in order to help customers identify relevant functionality. Today, these product packages contain many familiar products, but in time (over the next year to two years) some of these products will become more tightly coupled, and less like independent sets of products that a user will have to decipher and implement. This should also make it easier for a user to understand and buy.
The three groups, from bottom to top are The Server, The Analytics, and The Visuals. These names describe their primary logical functionality. Some products may share a presence in several product groups, however their predominant functionality will be represented in a grouping.
The Server gives our customers that all-important real-time reference platform upon which they can build applications, reports, and visualizations. Without this layer, they won't have a consistent way to use the information available to them. In other words: many versions of the truth (which is less than ideal).
The Analytics provide the server-based analytical functionality and products for The PI System. This now includes PI Analytics (logically) which are the familiar Performance Equations, Totalizers, Alarm, and RTSQC etc. Don't think about it in a deployment sense. That will change soon enough with PIANO and the roadmap. This is LOGICALLY where this functionality lives. It doesn't matter where the actual applications are getting installed. This is a way to help users think about our platform.
The Visuals are basically all the client products that talk to The PI System. Both Thin and Smart (or Rich) clients are represented in The Visuals. When we consider the Thin Clients, think of them as a "portal" (note the lower case "p") into the other PI System components… a way to see what's going on. Do NOT think of this as a "Portal" or allow users to get scared off by the name. We are not a portal vendor, but we make parts that can enhance their portal's functionality. Anything that we make that has a face on it will like here. So that may mean that certain visual elements may logically live here (like the RtReports Generator) even though they must get installed elsewhere.
All these groupings exist to help users understand the function of our products, not help them navigate product names or install the products we ship them.
The
Gathers, Organizes, Distributes, and Stores Data from Many Sources
Server
Structure / Asset Data
Real-Time Data
Custom Data
IT Data
Relational Data
Web Services
ERP / Maintenance 2 2 2

3. Value Now, Value Over Time
Value and Effort
Target
(time, value)
Batch Quality Monitoring
Effort
Asset Portfolio Optimization
Environmental Compliance
SQC
Six Sigma
Unlike expensive, “big bang” projects—such as ERP and EAM implementations—implementation of PI at your site happens rapidly, allowing your company to use of your data immediately. Because of this, the payback period on your investment is short.
As part of your Real-time Performance Management strategy, your company will establish PI System goals for your business. These equate to an overall Target, which is measurable in terms of the overall value you will obtain by implementing solutions built on the PI System real-time infrastructure platform.
Over time, you will build and implement solutions built upon the PI System, such as
Energy Management
Condition-based Maintenance (CBM)
Operational Equipment Effectiveness (OEE)
Six Sigma
SQC
Environmental Compliance
Asset Portfolio Compliance
Batch Quality Monitoring
Value Continues to Result as More is Built on the Platform!
OEE
Condition Based Maintenance
Energy Management
Time {
Initial
Software
Investment
Value
with PI
Incremental effort yields sustainable gains

4. Value Now, Value Over Time
New Value Targets will Appear
New Target
Value and Effort
Target
(time, value)
Effort
During your course of implementation or during the lifespan of your deployment of The PI System, new targets may appear. But, because you are building your value incrementally, you can change course without scrapping all your work or upgrading all your existing work at once.
To meet your new goal, you can build up new initiatives or change existing ones so that the value of The PI System continues to pay off.
Time {
Initial
Software
Investment
Value
with PI
Reach the New Target Incrementally

5. Intrinsic Values of the PI System
Your real-time data is:
Scalable (asset, site and enterprise)
Available (independent of system faults)
Reliable (believable, auditable)
Extensible (open feature set)
Secure (multiple security models)
Interoperable (open to multiple systems)
Scalability
Scalability refers to the capacity of the PI system to continue to perform and deliver value to the end user as the need for real time data processing expands by several orders of magnitude across an asset, a site and an enterprise.
Availability
Availability is defined by the capacity of the PI system and PI infrastructure to remain available to end users, to applications, across the enterprise and independent of faults that happen on the network, the computer platform, etc.
Reliability
Reliability means the capacity of the OSIsoft products to be dependable and the results delivered by PI believable and auditable.
Extensibility
Being extensible pertains to catering to the need that customer and partners have to extend the feature set and functionality of the PI System in an effective way.
Security
Security has evolved from a narrow access control focus to much broader topic. It now encompasses, in addition to application resident access control, such concepts as: integrated security with a common directory (Active Directory), more complex security models that are role based, pass through security concepts for system to system interactions, intrusion detection, forensics (ability to investigate security failure by looking at trace evidences), etc.
Interoperability, a.k.a Fit in place
This intrinsic value of Interoperability is more subtle and never forgotten by customers but under played by OSIsoft. The capability of a PI System to fit in place is essential to supporting scalability and enabling value for customers. Stronger focus on interfaces, smart connectors and pro active development of interfaces that can pull through business should be discussed.

6. Aspects of The PI System
Cross Functional Aspects of the PI System
The
Deliver Information for Meaningful, Better Business Decisions
Visuals
The
Turn Valuable Real-Time Data Into Actionable Information
Scalable
Available
Reliable
Extensible
Secure
Interoperable
Analytics
Across the PI System, we are constantly working to make our products scalable, available, etc.
The
Gathers, Organizes, Distributes, and Stores Data from Many Sources
Server 6 6 6

7. System Architecture Centralized Information PI Server LIMS Systems
Business
Systems
Portal
Servers
Firewall
Monitoring
Systems PI
Interface(s)
(Buffered)
Sensor PI
Server
Application
Servers
Client Apps
LIMS Systems
Data starts out at the source on the left… the sensor. It will be taking an analog (or maybe digital) signal and bringing it through some sort of acquisition system, to the PI System, all the way up to the corporate infrastructure. This site data is usually intimately tied to that site’s naming conventions (often produced by the control system implementer) and that site’s attributes. When people in the central office want to know something about the site, they either need help, or they need to understand that site’s operations intimately.
SCADA
Smart Devices
More Client Apps
Manual Data
PLC / Instrument
Systems
Maintenance Systems
Plant level app servers
Wireless
Mobile
Clients 7

8. The Server The Analytics The Visuals
PI Archive
PI AF
Real-Time Interfaces
Data Access
MCN Health Monitor
This is a one-level drill down into The PI System groupings. Today it lists the major products that live within the groupings. Keep in mind that this is going to morph as the roadmap begins to influence our platform. Ultimately the minor product names will disappear and will be replaced by functionality descriptions. People happen to know that PI does time-series data storage, but they don't know that it does a lot of other things. This is important to keep in mind when talking to an existing customer so that we can grow beyond providing them just a data historian.
Actual packages for sale are comprised of the products listed, but may not necessarily include all of them. For instance, even though RtReports is shown as part of The Analytics, it will not be always included in each edition of The Analytics that you might be selling, but it is still a member of that family. Same goes for RLINK and all the Smart Connectors we will make when it comes to The Server.
It is very important to notice that The Visuals have two major functional groups: Smart Clients and Thin Clients. There is more than one type of Thin Client we offer. One fits in Microsoft's SharePoint (we make Web Parts for this), another fits is SAP's Enterprise Portal (we make iViews for this). It is not correct to call the SAP offering "Web Parts for SAP" or the Microsoft offering "iViews for SharePoint" so… just don't do it. ;-)
Just like the Microsoft Office Family of products, there are different editions for sale, but if a potential customer wanted to know what product line FrontPage or Project lived in, Microsoft would say "Microsoft Office" even though you wouldn't necessarily buy those in the Microsoft Office 2003 Standard Edition product package. 8 8 8

9. The Server The Analytics The Visuals
PI Archive
ACE
PI AF
PI Analytics
Real-Time Interfaces
RtReports
Data Access
PI Notifications
MCN Health Monitor
Sigmafine
This is a one-level drill down into The PI System groupings. Today it lists the major products that live within the groupings. Keep in mind that this is going to morph as the roadmap begins to influence our platform. Ultimately the minor product names will disappear and will be replaced by functionality descriptions. People happen to know that PI does time-series data storage, but they don't know that it does a lot of other things. This is important to keep in mind when talking to an existing customer so that we can grow beyond providing them just a data historian.
Actual packages for sale are comprised of the products listed, but may not necessarily include all of them. For instance, even though RtReports is shown as part of The Analytics, it will not be always included in each edition of The Analytics that you might be selling, but it is still a member of that family. Same goes for RLINK and all the Smart Connectors we will make when it comes to The Server.
It is very important to notice that The Visuals have two major functional groups: Smart Clients and Thin Clients. There is more than one type of Thin Client we offer. One fits in Microsoft's SharePoint (we make Web Parts for this), another fits is SAP's Enterprise Portal (we make iViews for this). It is not correct to call the SAP offering "Web Parts for SAP" or the Microsoft offering "iViews for SharePoint" so… just don't do it. ;-)
Just like the Microsoft Office Family of products, there are different editions for sale, but if a potential customer wanted to know what product line FrontPage or Project lived in, Microsoft would say "Microsoft Office" even though you wouldn't necessarily buy those in the Microsoft Office 2003 Standard Edition product package. 9 9 9

10. The Server The Analytics The Visuals
Smart Clients
PI Archive
ACE
PI ProcessBook
PI AF
PI Analytics
PI DataLink
Real-Time Interfaces
RtReports
PI Activeview
Data Access
PI Notifications
PI BatchView
MCN Health Monitor
Sigmafine
Thin Clients
RtWebParts
This is a one-level drill down into The PI System groupings. Today it lists the major products that live within the groupings. Keep in mind that this is going to morph as the roadmap begins to influence our platform. Ultimately the minor product names will disappear and will be replaced by functionality descriptions. People happen to know that PI does time-series data storage, but they don't know that it does a lot of other things. This is important to keep in mind when talking to an existing customer so that we can grow beyond providing them just a data historian.
Actual packages for sale are comprised of the products listed, but may not necessarily include all of them. For instance, even though RtReports is shown as part of The Analytics, it will not be always included in each edition of The Analytics that you might be selling, but it is still a member of that family. Same goes for RLINK and all the Smart Connectors we will make when it comes to The Server.
It is very important to notice that The Visuals have two major functional groups: Smart Clients and Thin Clients. There is more than one type of Thin Client we offer. One fits in Microsoft's SharePoint (we make Web Parts for this), another fits is SAP's Enterprise Portal (we make iViews for this). It is not correct to call the SAP offering "Web Parts for SAP" or the Microsoft offering "iViews for SharePoint" so… just don't do it. ;-)
Just like the Microsoft Office Family of products, there are different editions for sale, but if a potential customer wanted to know what product line FrontPage or Project lived in, Microsoft would say "Microsoft Office" even though you wouldn't necessarily buy those in the Microsoft Office 2003 Standard Edition product package.
RtPortal iViews
RtReports Clients 10 10 10

11. The Server The Analytics The Visuals
Smart Clients
PI Archive
ACE
PI ProcessBook
PI AF
PI Analytics
PI DataLink
Real-Time Interfaces
RtReports
PI Activeview
Data Access
PI Notifications
PI BatchView
MCN Health Monitor
Sigmafine
Thin Clients
RtWebParts
This is a one-level drill down into The PI System groupings. Today it lists the major products that live within the groupings. Keep in mind that this is going to morph as the roadmap begins to influence our platform. Ultimately the minor product names will disappear and will be replaced by functionality descriptions. People happen to know that PI does time-series data storage, but they don't know that it does a lot of other things. This is important to keep in mind when talking to an existing customer so that we can grow beyond providing them just a data historian.
Actual packages for sale are comprised of the products listed, but may not necessarily include all of them. For instance, even though RtReports is shown as part of The Analytics, it will not be always included in each edition of The Analytics that you might be selling, but it is still a member of that family. Same goes for RLINK and all the Smart Connectors we will make when it comes to The Server.
It is very important to notice that The Visuals have two major functional groups: Smart Clients and Thin Clients. There is more than one type of Thin Client we offer. One fits in Microsoft's SharePoint (we make Web Parts for this), another fits is SAP's Enterprise Portal (we make iViews for this). It is not correct to call the SAP offering "Web Parts for SAP" or the Microsoft offering "iViews for SharePoint" so… just don't do it. ;-)
Just like the Microsoft Office Family of products, there are different editions for sale, but if a potential customer wanted to know what product line FrontPage or Project lived in, Microsoft would say "Microsoft Office" even though you wouldn't necessarily buy those in the Microsoft Office 2003 Standard Edition product package.
RtPortal iViews
RtReports Clients 11 11 11

12. The PI Server Today 4/21/20174/21/2017 The Server 1212 PI Archive The
Analytics
The Visuals
Data Access
RtBaseline Services for Thin Clients
OLEDB
ODBC
OPC / HDA
Module Database
PI Archive
Failover / Management Services
MCN HealthMonitor & System Management Tools
AF 2.0
Data Storage, Management Services, Audit Trail, Batch
Custom Programming
RLINK
At Platform Release 2 we introduce the concept of a “Data Directory” instead of the Module Database and Analysis Framework. The Data Directory is a collection of many features of the platform. The PI AF 2.0 component is the first output of the Foundation development initiative. PI AF 2.0 is an entirely new infrastructure compared to PI AF 1.x. It now requires some version of SQL Server 2005.
The Data Directory and the initial delivery of PI AF 2.0 is the beginning of having a more generalized way of finding and accessing data in The PI System. The new database being introduced will do the Analysis Framework type functions as well as some of the Module Database functions. For backward compatibility reasons, the Module Database is still shipped with PI and old AF applications will still work.
More will happen with the Data Directory in PR 3 and beyond.
Also, further work will be going on in the Failover and High Availability functions of The Server. Also, features like future data, and better security integration, and data quality are being added incrementally to all the functions listed in the diagram.
Smart Connectors
Smart Connectors represent all of OSIsoft's connectivity technology. The primary members are Interfaces (which connect to systems that usually provide streaming or real-time data), Asset Connectors (an emerging class of connectors that synchronize hierarchal information about assets) and Business Gateways (currently the closest product we have for this are the RLINK Gateways, but others will come).
The goal of any member of the Smart Connector family is to connect to a foreign system and pull as much information into The PI System as possible with as little effort as possible. Development is ongoing to ease the configuration burden experienced by a user. Technologies like AutoPointSync, the Interface Configuration Utility, and new tools will have to be further automated or created in order to achieve the goal of near-zero configuration.
In the future, many target devices may require an Interface, an Asset Connector, and a Business Gateway. The goal is to be able to deploy all these technologies seamlessly if required. The first Asset Connectors will be introduced at the time that PR 2 is released. These Asset Connectors will have a similar design to our Real-time Interfaces in that they will have a core of technology like UNIINT that keeps the PI AF 2.0 database in sync with a foreign system. The work to create an Asset Connector will become fairly well known, just like the work needed to create a new Real-time Interface.
Business Gateways will also start to be introduced around the time frame of PR 2. They will take on some of the tasks of RLINK. They may require calculations, asset data, and real-time data in order to perform their tasks. The idea being that we develop a meaningful connection to business systems or maintenance systems without duplicating our data up in those systems.
Even though the PI Analytics are not listed in The Server, the PIANO project will be in its first release at Platform Release 2. This release will allow some of the traditional PI Analytics to be hosted in another server (grouped with The Analytics).
In the Data Access family, there will be some changes in the PI OLEDB provider to make it faster and more useful for large Extract, Transform, and Load operations. Also, an OLEDB provider will be introduced for PI AF 2.0.
Real-time Interfaces
Structure / Asset Data
ERP / Maintenance RLINK SERVERS & GATEWAYS SAP / JD EDWARDS / PEOPLESOFT MAXIMO / INDUS / MIMOSA
Real-Time Data DCS / PLC / SCADA / OPC HISTORIANS / INTERFACES COM CONNECTORS
Custom Data APIs / SDKs
IT Data IT MONITOR
Relational Data OLEDB / ODBC ORACLE / SQL
Web Services SOA / EXTERNAL DATA LEGACY APPS
Copyright © 2007 OSIsoft, Inc. All Rights Reserved
1212 12 12

13. The Analytics within the PI System Today
4/21/2017
The Analytics within the PI System Today
The
Analytics
The
Server
The Visuals
Advanced Computing Engine (ACE)
PI Notifications
RtReports Compliance & Standard
Enterprise Services
PI Analytics (Performance Equations, Totalizers, Alarm, RTSQC)
Members of The Analytics group of products are, in some cases, unrelated to each other, but very related to the components of The Server. These products or functions are all analytical in nature (RtReports does quite a bit of analysis prior to generating a report). These are usually centrally installed applications that post results back to The Server. Some may have dedicated clients for use and configuration purposes, however the actual client components will live in The Visuals. The parts that live in The Visuals are the components that render the analytics or reports, not the applications doing the calculation work.
The Analytics logically house the PI Analytics group, such as Performance Equations, Totalizers, RTSQC, and Alarm live in this product group now. This is largely due to the PIANO development initiative and the product roadmap. These functions will be able to reference more than one PI Server and also more than just PI data. Therefore, binding them to PI really isn't correct anymore even though this will be haw these elements are delivered in the next release.
The other major products in The Analytics are fairly self explanatory. They need a home since they aren't quite clients, and aren't quite servers, or are a blend of everything. So a product that has several parts may get split up between different families (logically speaking). This is especially true if they have different licensing bases (such as RtReports and Sigmafine). If it's split up, the part that crunches goes in The Analytics. The part with a face on it lives in The Visuals. If it's % about the crunching, then it definitely lives in The Analytics! 13 13

14. The Visuals Smart Clients Thin Clients ProcessBook RtWebParts DataLink
4/21/2017
The
Visuals
The
Server
The Analytics
Smart Clients
Thin Clients
BatchView SQC Client AF Modeler Add-in
RtActiveView RtGauge RtGraphic RtTable RtTagSearch
RtTreeView RtTrend RtTable RtXYPlot
ProcessBook
RtWebParts
MS Excel Add-ins:
BatchView AF
RtActiveView RtGauge RtGraphic RtTable RtTagSearch
RtTreeView RtTrend RtTable RtXYPlot RLINK iViews
DataLink
iViews
Members of The Analytics group of products are, in some cases, unrelated to each other, but very related to the components of The Server. These products or functions are all analytical in nature (RtReports does quite a bit of analysis prior to generating a report). These are usually centrally installed applications that post results back to The Server. Some may have dedicated clients for use and configuration purposes, however the actual client components will live in The Visuals. The parts that live in The Visuals are the components that render the analytics or reports, not the applications doing the calculation work.
The Analytics logically house the PI Analytics group, such as Performance Equations, Totalizers, RTSQC, and Alarm live in this product group now. This is largely due to the PIANO development initiative and the product roadmap. These functions will be able to reference more than one PI Server and also more than just PI data. Therefore, binding them to PI really isn't correct anymore even though this will be haw these elements are delivered in the next release.
The other major products in The Analytics are fairly self explanatory. They need a home since they aren't quite clients, and aren't quite servers, or are a blend of everything. So a product that has several parts may get split up between different families (logically speaking). This is especially true if they have different licensing bases (such as RtReports and Sigmafine). If it's split up, the part that crunches goes in The Analytics. The part with a face on it lives in The Visuals. If it's % about the crunching, then it definitely lives in The Analytics!
DataLink for Excel Services
RtReports Generator
RtReports Editor
ActiveView
Other Thin Clients
Performance Improvement
Operational Visibility
Knowledge Management
Product Quality
Lean Manufacturing
Asset Management
Situational Awareness 14 14

15. The Visuals Smart Clients Thin Clients The The Analytics Server
4/21/2017
The
Visuals
The
Server
The Analytics
Smart Clients
Thin Clients
Visualization
Root Cause Analysis
Planning
Content Authoring
Structured Portal Environment Allows
Users to Build Displays Without IT Training
Seamless Interaction
Between Smart Clients & Portal Environment
Members of The Analytics group of products are, in some cases, unrelated to each other, but very related to the components of The Server. These products or functions are all analytical in nature (RtReports does quite a bit of analysis prior to generating a report). These are usually centrally installed applications that post results back to The Server. Some may have dedicated clients for use and configuration purposes, however the actual client components will live in The Visuals. The parts that live in The Visuals are the components that render the analytics or reports, not the applications doing the calculation work.
The Analytics logically house the PI Analytics group, such as Performance Equations, Totalizers, RTSQC, and Alarm live in this product group now. This is largely due to the PIANO development initiative and the product roadmap. These functions will be able to reference more than one PI Server and also more than just PI data. Therefore, binding them to PI really isn't correct anymore even though this will be haw these elements are delivered in the next release.
The other major products in The Analytics are fairly self explanatory. They need a home since they aren't quite clients, and aren't quite servers, or are a blend of everything. So a product that has several parts may get split up between different families (logically speaking). This is especially true if they have different licensing bases (such as RtReports and Sigmafine). If it's split up, the part that crunches goes in The Analytics. The part with a face on it lives in The Visuals. If it's % about the crunching, then it definitely lives in The Analytics!
Performance Improvement
Operational Visibility
Knowledge Management
Product Quality
Lean Manufacturing
Asset Management
Situational Awareness 15 15

16. PI Server

17. Point Count
PR1 ( ) x64
PR2+ (3.5) x64
3.4 (UC 2003)
x86
x64
PI 3.3 SR2 ,
Release Dates:
3.3 March 2001
3.3 SR2 November 2002
3.4 January 2004
November 2004
October 2005
Jul 2006
x64 October 2006
December 2006 – PR1
April 2008 – PR1 SP1 17

18. In Other Words…
>100x

19. Startup Time

20. Archiving Rate

21. Snapshot Rate

22. Archive Query Rate

23. Motivations: Why High Availability?
Network Failure
Administrative/User Failure
Upgrades and Migrations
Hardware
Operating System Software (Patches)
Applications Software
Software Failure – although Microsoft, Emerson, and OSIsoft test our software – software systems in a time of change can fail.
Hardware Failure – servers have become remarkably more reliable but failures can still happen
Network Failure – networks have also become more reliable, but also much more complex. These also provide some risk.
Administrative/User Failure – when putting in new firewalls, or performing software, hardware, or networking upgrades, or performing administrative procedures, people make mistakes and those mistakes can result in data and availability loss.
Operating System upgrades – patches for security are increasingly sensitive – also new versions of the operating system software are do this year and High Availability will assure that data and users will not be interrputed as this is tested and implemented
Emerson/OSIsoft upgrades – Emerson and OSIsoft are continuously improving their products. Get the newest DeltaV and PI features without interruptions to data or users.
Hardware upgrades / migrations – migrating to a new platform can be done without data or access loss.

24. The Business Case for HA
How much downtime can you afford?
365 Days X 24 Hours = 8760 Hours/year
99.7% uptime = 26.3 hours of downtime/year
99.9% uptime = 8.7 hours of downtime/year

25. The Value of High Availability today
More uptime during server outages
Planned
Unplanned
Automatic failover for data consumers
Automatically publish changes from a primary to secondary servers
Multiple configurations possible

26. PI SDK Discovery, Failover, Failback, Load Distribution Services
PI HA Architecture
Clients: ProcessBook, DataLink, BatchView, RtWebParts, Custom Application…
System Management Tools
AppServers: AF, RtBLS, RtReports, ACE
PI SDK Discovery, Failover, Failback, Load Distribution Services
API Buffering
PI Collective
Secondary
PI Server
Primary
PI Server
Secondary
PI Server(s)
Configuration
Changes
Configuration
Changes
API Buffering Services
PI Interfaces
API Buffering Services
PI Interfaces
Failover
Mechanisms

27. PI HA Summary Features Value
Synchronization of PI Server configuration
Transparent client failover, simple load balancing
Fanning of real-time data to multiple servers
Value
High Availability to your PI System
Peace of mind for Administrators
Direct support for existing PI Clients
Simple, scalable and flexible architecture
Direct support
No changes to displays or spreadsheets. New SDK provides Failover even with older apps
Scalable architecture:
Support for systems of all sizes
No specialized hardware requirement
Geographic Availability
Disaster Recovery

28. PI Server Roadmap PI 3.4.375.80 – PR1 Service Pack 1 – Released
PI Server with Windows Integrated Security
Presented by Chuck Muraski later today

29. T1 T2 PI Archive T3 T4 PI Server:
View of all your real-time process data wherever you need it
Highly scalable
High performance
Rapid deployment and easy to configure and maintain
Become an essential tool to process engineers around the globe
Points database centric: a list of all the points collected from sensors throughout your process. As # of points increases, and the people viewing the data have less expertise in the PI server, a new problem arises: What data should I look at? How do I find my data? These are problems that OSIsoft has been addressing of late. One of the ways we are doing this is by adding new ways to get at your real-time data that has been stored in PI for years…
How many tags can you memorize? Does your job description call for memorizing tag names?
How consistent is your tag naming convention?
What about turnover and new people learning the tags?
Next: Context data. T4

30. Physical Context T1 T2 PI Archive T3 T4
Physical context or logical context in the case of a plant, line or grouping or abstraction of assets.
Accomplished with the MDB and AF 2.0.
Show Oil rig and then tank and talk about how the question is more in the context of the plant than the database.
Also mention that if you have two oil rigs, you can build displays or calculations to look at an oil rig generically and then apply the context of which oil rig you are concerned with to see the right calculation or visualization. T4

31. Temporal Context T1 T2 PI Archive T3 T4
Click - Batch, startup, shift (lower picture).
Click – Allows you to frame the real time data around the event or batch
Click – Gives you a view into the relevant time period of your real-time data
Accomplished with the Batch Database today and Event Frame Database in the future. T4

32. PI Module Database – Asset Hierarchy
Division Level
Asset
District Level
Equipment Class
Individual Equipment

33. PI Module Database – Attributes
Alias
Configuration
Individual Equipment
Aliases
Property
Configuration
Properties

34. PI Module Database – Asset Relative Displays

35. PI Module Database – Asset Relative Displays

36. PI DataLink – Asset Relative Reports

37. PI DataLink – Asset Relative Reports

38. Next Generation of Asset Database
PI AF 2.0 succeeds PI Module Database
OSIsoft is committed to moving customers and partners forward to AF 2.0…
…While preserving value of existing applications

39. Features Module Database AF 2.0 Hierarchical  Static values PI Points
Versioning
Event Pipes *
Templates
Formulae
Units of Measure
External Data
Connectivity
Event Pipes are not yet implemented on the AF 2.0 database. However, there is a mechanism for a client to poll for the updates to the AF database since a certain time. It is not subscription based and does not provide events (needs to be polled). We are planning on implementing Event Pipes in AF.
Templates is very powerful concept. Give compressor example. Compressor template with various attributes like capacity, maintenance, rpm, pressure. Now build compressor #1 from the template and voila you get those attributes ready with default values you can change.
More advantages:
Element by element security
More scalable
Searchable
More details on this in other talks at the conference.

40. Pipeline Example Natural Gas Compressor RPM Fuel consumption rate Tags
Exhaust temperature
Oxygen sensor
Flow in
Flow out
Pressure differential
Last service date
Service type
Mfr documentation
Tags
Calculations
UOM
Pipeline has compressor stations along the way to maintain pressure and hence flow of the gas. It may have properties like these.
SQL Query

41. Pipeline Example Build compressor template Create attribute templates
Set default values
Create instances of template
Compressor1
Compressor2
CompressorN
Adjust data streams
With those properties, you first build a template. This represent the class of equipment, it is not an instance of the equipment. This will not necessarily point to any data, but will have placeholders for them. You then create the attribute templates. When you create instances of the compressor to represent the compressors on your pipeline, you’ll name them and then make sure the data stream point to the correct PI tags.

42. Template Example Connect compressors Model/Connectivity
You could connect them up in a model.

43. Now you can… Locate data without needing tag names
Build just one display and show each compressor
Include non-PI data in reports
Estimate next maintenance
Calculate fuel consumed
Calculate losses find leaks
Find which compressors have low pressure differentials.
And so on…
You can do all these things and more. The ease of building reports or displays once and applying to each instance is quite appealing. It can save time and money now and as you need to maintain displays. Also helps you to keep a common report or display.

44. AF 2.0 Assists Users Search your data more intuitively
Access more than PI data
Define generic calculations
Create display and report Templates
Scale to the Enterprise
DataSheet:
The AF from OSIsoft® is a flexible product that enables an organization to define a consistent representation of its assets and use these assets in simple to complex analyses that yield critical and actionable information. Through its intuitive user interface, AF allows an application designer to
1. Identify the components or elements that make up a process;
2. Associate data—be it real-time or relational data—with these elements;
3. Organize elements into logical groups (e.g., asset templates) and specify
relationships;
4. Apply calculations, rules, or solvers; and
5. Determine how results are to be displayed.
Once defined, individual assets can be organized and regrouped, without limitations, to drive meaningful analyses and enable rich presentation of data. Because it is configuration-driven rather than programmatically-driven, the Analysis Framework
allows users to accomplish all of these tasks easily and rapidly.

45. AF Features Assets can be: Attributes can be:
Categorized
Connected
Versioned
Hierarchical and Linked
From an existing Asset Model – “Smart Connectors”
Attributes can be:
PI Points
From Relational Databases
Calculated
Inherited from a Template
XML Compatible (import and export)

46. PI AF 2.0 – New Features Performance and scalability
Millions of assets mapping multi-millions of PI points
Multiple PI Server Support
Object History (versioning)
Keeps track of configuration changes over time
Hierarchical Attributes and Categories
Advanced Attributes Types
Files, Arrays, Tables
Enumeration Sets
Object-Level Windows Integrated Security

47. PI AF 2.0 Benefits Visibility Single Version of the Truth
Information about data, assets and models is available to visualization suite of tools
Single Version of the Truth
Creates unified assets and relationships
Assets and models become basis for consistent visualization and analyses
Communication & Collaboration
People use the same asset and model reference for their visualization and analyses
Better understanding, communication and collaboration
Visibility (enterprise-wide data, assets, analyses)
Knowledge management
Information about data, assets, models and analysis is available to visualization suite of tools
Single Version of the Truth: centralization of knowledge that is easily and logically stored for use by the entire enterprise
Easily share in the company’s knowledge as it pertains to a particular asset and in the context of a particular role
People use the same asset and model reference for their visualization and analyses
The advantage of such an open, flexible infrastructure is that it enables companies to configure a solution application—which is a set of plug-ins such as complex calculations, PI Data Reference, Mass Balance Analysis, Gross Error Analysis, Units of Measure conversion, etc—in order to analyze virtually any part of the operation.

48. PI AF 2.0 Benefits Search and Find information Knowledge Management
Expose a more intuitive, equipment centric view of our data so everybody in the corporation can quickly access the data relevant to their jobs
Knowledge Management
Synchronize and eliminate disparate assets, models and analysis to improve traceability and accountability
Better understanding, communication and collaboration
Visibility (enterprise-wide data, assets, analyses)
Knowledge management
Information about data, assets, models and analysis is available to visualization suite of tools
Single Version of the Truth: centralization of knowledge that is easily and logically stored for use by the entire enterprise
Easily share in the company’s knowledge as it pertains to a particular asset and in the context of a particular role
People use the same asset and model reference for their visualization and analyses
The advantage of such an open, flexible infrastructure is that it enables companies to configure a solution application—which is a set of plug-ins such as complex calculations, PI Data Reference, Mass Balance Analysis, Gross Error Analysis, Units of Measure conversion, etc—in order to analyze virtually any part of the operation.

49. PI AF 2.0 in the PI System PI AF 2.0
Data structured and organized by asset
Spans multiple PI Systems
Incorporates non time series data
PI AF 2.0
Asset-Equipment Centric access to the Data
Non Time Series Data Sources …
SCM
RDBMS
MES
Honeywell
ABB …
Time Series Data Sources
OPC
Interface
XML
Modbus
SNMP
Generic Interfaces
Delta-V
Rockwell
Vendor Specific Interfaces
PI Server 2
Other
Honeywell
ABB …
Time Series Data Sources
OPC
Interface
XML
Modbus
SNMP
Generic Interfaces
Delta-V
Rockwell
Vendor Specific Interfaces
PI Server 1
Other

50. PI AF 2.0 Architecture Scenario 1 Scenario 2 AF clients AF clients
AF Server
AF Server
AF Database
Scenario 1
Scenario 2

51. Features: PI System Explorer
View the structure of the database
Templates, Elements, Relationships, Tables, Notifications
Create new content
All AF objects
Configure items
assign data references
Create structures
Hierarchies, models

52. Features: ProcessBook Modeler
Provides a visual representation of AF
The AF Toolbar provides a mechanism for setting the start and end dates for case studies.
The Case Runner Control keeps track of the status of a running model, indicating the approximate time remaining.
The Model section provides a complete representation of the model. The various colors indicate the status of each component. Each meter contains information specific to the measurement being taken.
The AF Explorer Window provides access to General information, elements, attributes and ports related to individual modules.

53. Features: Excel Add-in
Manage configuration
Specify data references for attributes that will override the template definition
Report information
Create report templates similar
to using DataLink
Retrieve information regarding model connectivity and context
The Excel Add-In comprises of 2 parts – data and connectivity retrieval, and element creation and configuration.
With the AF element configuration the user can specify the data references for attributes that will override the template definition.
The data retrieval allows the user to create report templates in Excel similar to using Data link, but it has the knowledge of the model and connectivity and context.

54. Features: AF SDK Comprehensive Managed Code Class Library
.Net based
Access to AF Database
Common UI Components
Rich library of controls
Extensive User Guide and Programmer Reference
Tutorial Guiding Developer Through AF Programming Experience

55. PI AF 2.0 Advantages
Allows organization of all of your data -millions of points - in a meaningful way
According to an equipment model
According to the way you would use it in calculations
Create relationships between collections or organizations (e.g., assets)
Allows you to bring in relational and complex data into the same context as your real-time data

56. PI AF 2.0 Advantages
Custom domain knowledge can be re-used in new analyses because it’s in one central place (not in an Excel spreadsheet)
Replaces programming with configuration
Protects your investments in your displays, reports, and business logic

57. MDB to AF Give me a Module MDB AF 2.0 Give me an Element
Possible that MDB to AF 2.0 is many to one.
Possible that MDB could see all of AF 2.0
Discuss how batch will work too

58. AF Roadmap AF 2.0 is released MDB to AF migration and connection
Allows AF to be system of record
Allows MDB calls to read AF Elements as they map to Modules
Allows ACE to work against AF Element contexts
Requires Windows Security PI Server
Platform Release 2 (PR2) – “AF Unification”
Clients based on AF rather than MDB
ACE will read AF Element contexts natively
AF High Availability

59. PI Notifications (Bridge) & PI Analytics (PR 3)
RtAlerts -> PI Notifications
ACE, Analysis Rules -> Best practices for PI Analytics
Single metadata and structure basis
Uses any data referenced in The PI System
Smart Connectors
DataLink for Excel Services
PI Security
PI Batch
PI Reporting

60. What is PI Notifications?
A scalable, highly available, notification and alarming infrastructure
Why infrastructure?
Extensible
Users can customize notification rules and scheduling
Integral part of the PI System
Real-time event delivery
PI System Client integration
Consistent with Enterprise Message of PI 60

61. PI Notifications: The Value
Alerting “out of band” – anywhere, anytime
Escalation
Notification history (including acknowledgements and other actions) maintained forever
No programming required
Extensible delivery channels that can integrate with 3rd party systems and applications

62. PI Notifications Topology

63. PI Notifications Business Logic Notification Emission Analysis Rule
Time
Target
Configuration
Notification Emission
Trigger
Contacts
Content
Delivery
Acknowledgement

64. The PI Notification Process
PI Notification Service collects inputs and evaluates the Trigger condition(s).
Subscriptions are evaluated for the correct Contact(s).
Content is added to the Notification.
Subscriptions are filled via delivery channels.
Subscribers can acknowledge the Notification.

65. Notifications
Once a Trigger condition is satisfied, a historical record is created for that condition.
Only the Trigger is required to create a PI Notification - Content and Subscriptions are optional.

66. PI System Explorer Integration

67. PI ProcessBook Integration

68. PI DataLink Integration

69. Notifications Roadmap
PI Notifications 1.0 is released
PI Notifications enhancements planned
Scale up performance
MyPI Enhancements
Searching
Grouping
Snooze Notification Rule
Internationalization

70. PI ACE – Consistent Performance Metrics
Scalable application development environment
One, consistent calculation can be used for all similar processes or equipment
Integrates access to data and functionality from other systems

71. Continuous Data

72. Batch

73. Day or Shift

74. Event
Startups
Downtime
Shutdowns
Excursions/incidents

75. What is an Event Frame? Start time End time Name Associated data
Equipment
Tags
Comments
Event Frames
2008Aug06-ABC123
New data type defined by a name, start time, end time and optionally accompanied by supporting data
Start time
End time
Name
Associated data (Equipment, Tags, Comments, other Event Frames)

76. Batch and Event Analysis
Every process has some kind of repeatable event
Batch process
Am I making it the same way as yesterday?
How do I make the best product?
How can I reduce cycle times?
Continuous process
What causes downtime?
Can I improve my startup or grade change times?
How can I track and study environmental excursions?
Discrete process
What is my Overall Equipment Effectiveness?
What are my top 10 causes of downtime?
Study these events to improve your process. In some cases you want to repeat these events and in other cases you want to avoid or minimize their effects.

77. PI Batch Architecture PI EVT Interface PI Archive PI BaGen
Batch Exec. Systems
Rockwell RSBatch
Emerson DeltaV
Intellution iBatch
Honeywell Total Plant Batch
PI EVT Interface
PI Archive
PI Batch Generator
Interface
PI BaGen
Batches and Modules
Batches
PI Point Events
PI Server
Module Database
Batch Database
S88

78. PI Batch Architecture Analytics PI Archive Visualization Smart Clients
ProcessBook
Excel/Datalink
PI Archive
Analytics Server
RtReports
RtBaseline Services
Analytics
Visualization
Trust
PI Server
Module Database
Batch Database
Internet Explorer
Thin Clients
RtWebParts
RtReports

79. Batch and Event Frames Roadmap
PI Batch suite is released product
Event Frames work in progress

80. Thank You
Questions?