© 2019 AVEVA Group plc and its subsidiaries. All rights reserved.

Historian Innovation:
What’s New & What’s Coming

Session ID: MC-25

Presented by: Elliott Middleton, Director of Product Management

November 14th, 2019

© 2019 AVEVA Group plc and its subsidiaries. All rights reserved.

AVEVA Fiscal Year 2019 Highlights
ending March

System Platform Historian Standalone Historian

+23% +37% +47%

Across all regions

Across all tag counts

License + Subscription Orders

 How Customers Use Historian
Diagnostics/troubleshooting 71%

Regulatory reporting 46%

Identifying improvement opportunities 43%

Downtime tracking 29%

Production/cost accounting 27%

Energy consumption 19%

0% 25% 50% 75%

Multiple responses allowed
Source: Tech Validate
Name Changes

2017 (17.0) 2020 (20.0)

Wonderware Historian AVEVA Historian
(formerly Wonderware)

Historian Client Historian Client Desktop

Historian Insight Historian Client Web

Client Licensing Options: Concurrent User

Perpetual FLEX
1 Historian Client Desktop
3
2 Historian Client Web
What Makes A Historian?

Presentation

Transformation

Preservation

Collection
Summarizing Analog Measurements
From Historian

Statistics Σ
x σ
ad hoc 9.0 - 2005

stored (configured) 10.0 - 2010

Outliers
highlight
ignore stored (automatic) 2017
Periodic Statistics

Example: Hourly averages for a day

Example Query: Hourly Data
select TagName, StartDateTime, EndDateTime, wwResolution, Average
from AnalogSummaryHistory
where TagName = 'M21.FlowIn'
and StartDateTime >= '2019-01-15 0:00'
and EndDateTime <= '2019-01-16 0:00'
and wwResolution = 3600000

Each row has statistics

for an hour
What About Conditional Statistics?

Average flow rate while valve is open

Custom Stored Procedure: Conditional Statistics
create function dbo.wwkbSliceBy(@StartTime datetime2, @EndTime datetime2, @SliceByTag nvarcha
returns @SliceResults table
(
StartDate datetime2,
EndDate datetime2,
SliceBy sql_variant,
TagName nvarchar(128),
Average float,
Duration int
)
as
begin
declare @Slices table(DateTime datetime2, vValue sql_variant, wwResolution int, Quality int)
insert @Slices(DateTime, vValue, wwResolution, Quality)
select DateTime, vValue, wwResolution, Quality
from History
where DateTime > @StartTime and DateTime < @EndTime
and TagName = @SliceByTag
and wwRetrievalMode='delta'
New In Update 3: “Slice By” Query
select TagName, StartDateTime, EndDateTime, wwResolution, SliceByValue, Average
from AnalogSummaryHistory
where TagName = 'M21.FlowIn'
and StartDateTime >= '2019-01-15 0:00' Row for each “SliceBy”
and EndDateTime <= '2019-01-16 0:00' value change
and SliceBy='M21.ValveIn'

Each row has

statistics for
corresponding
time
What About By Batch Phase?

Peak temperature by batch phase

Peak Temperature By Batch Phase
select TagName, StartDateTime, EndDateTime, wwResolution, SliceByValue, Maximum
from AnalogSummaryHistory
where TagName = 'M21.Temp'
and StartDateTime >= '2019-01-15 0:00' Row for each “SliceBy”
and EndDateTime <= '2019-01-16 0:00' value change
and SliceBy='M21.Phase'
What About By Batch Phase?

Peak temperature by batch

Peak Temperature By Batch
select TagName, StartDateTime, EndDateTime, wwResolution, SliceByValue, Maximum
from AnalogSummaryHistory
where TagName = 'M21.Temp'
and StartDateTime >= '2019-01-15 0:00' Row for each BatchID
and EndDateTime <= '2019-01-16 0:00' change
and SliceBy='M21.Batch'
Bounding Values Query is for this period
Boundary Values
Excluded Included
select TagName, StartDateTime, EndDateTime, select TagName, StartDateTime, EndDateTime,
wwResolution, SliceByValue, Maximum wwResolution, SliceByValue, Maximum
from AnalogSummaryHistory from AnalogSummaryHistory
where TagName = 'TI101.PV' where TagName = 'TI101.PV'
and StartDateTime >= '2019-01-15 7:00' and EndDateTime >= '2019-01-15 7:00'
and EndDateTime <= '2019-01-15 8:00' and StartDateTime <= '2019-01-15 8:00'
and SliceBy='M22.ValveIn' and SliceBy='M22.ValveIn'

Start is before “7:00”

End is after “8:00”

 Time Slicing Options

Historian Intelligence
• Simple, ad hoc queries • Pre-calculated for performance
• Single slicing criteria* • Multiple slicing criteria
(yield by batch, recipe, line, operator)
• Data within Historian
• Also external data
* REST supports multiple criteria
Easily Combine
 Dashboard Comparison
Insight Intelligence

• All data is time-series • Virtually any data

• Super-simple, self service • Leverages time-slicing engine
• Independent charts • Expert user enables self-service
• Basic layout • Enables chart interaction
 Alarm History Options
• Routine purge required
• Manageability • Reliable forwarding (SQL Server limitation)
(HCAL)
• Reliability could be better
Historian A2ALMDB WWALMDB WWALMDB

Alarm DB
HCAL
Logger
Application Server Application Server Application Server InTouch

2014 R2 2014 1.0 - 2002 7.11 - 2000

 Block-Based Alarm History

• Performance
Historian • Significantly higher storage rates
• Capacity limited by disk space, not insertion rate
HCAL

Application Server • Easier To Manage

• Archive/purge/restore using Windows Explorer
• No need to purge to sustain storage
2014 R2
Alarms In Context
Alarms In Context

Historian 2017 Update 1 Historian Client 2017 Update 2

Online Insight

Insight Mobile

© 2018 AVEVA Solutions Limited and its subsidiaries. All rights reserved.
Which Alarms?

Historian WWALMDB

Application Server InTouch

Control System

© 2018 AVEVA Solutions Limited and its subsidiaries. All rights reserved.
Which Alarms?

Historian

Application Server Detected Here InTouch

Local Publisher required
• Software included with Update 3
Control System Detected Here • Disabled until System Platform 2020

© 2018 AVEVA Solutions Limited and its subsidiaries. All rights reserved.
Which Alarms?

Historian 2020 Historian 2020

Application Server 2020 Detected Here InTouch

Will require Remote IDAS here

Control System Detected Here

© 2018 AVEVA Solutions Limited and its subsidiaries. All rights reserved.
External Alarms
Alarm Metrics
HiHi = 184.0

Hi = 179.0

How many alarms are active at this time?

What was the total time in “Hi” alarm

alarm?above the high limit?
A
B
 Model-based navigation (“breadcrumbs”) online Q4
on-premises ?
Grouped by Attribute, Tag, Object, Area
Tiered Historians 10.0 - 2010

• Make it easy to manage

• Make efficient use of network

9.0 - 2005
“I don’t know how many of you are
using tiered Historians, but you need to.
Just get it!”
-- Oil & Gas Customer
 * Historian 2017 Update 3 SP1
Historian Replication (N-Tier)
Limitation: Cannot re-summarize
“simple” to “simple”
“simple” to “1-hour”
“1-hour” to “1-hour”
“1-hour” to “1-day” Tier-3* Enterprise

Tier-2 * Region

Tier-1 Site
Excel Add-ins
Historian Client

• Ribbon bar • Broad, mature features (12)

• Add-in functions • 32-bit Excel

• Task pane • Raw data, summary statistics

• Import (no functions) • 64/32-bit “ClickOnce” Excel
Historian Insight
Historian 2020 Excel Add-in
• Tag Selection: Based on search
• Tag Details
• Meta data
• Static or function

• Single Value
• Point in time (including “now”)
• Static or function

• Periodic Values (Analog Summary)

• Time In State (State Summary)
Excel 2016/365 Add-in
Both 32-bit & 64-bit

U3
Historian Retrieval Bottlenecks
Example Historian
Client
With 20 tags, reduces
SQL Server
row count to ~1/20 SQL
Throttles results sets > ~10K rows Insight
Server
May reduce rows with “WideHistory”
OLEDB REST SDK
Example
SQL: 20 seconds @ 10% CPU “High-Level Retrieval”
Direct: 2 seconds @ 100% CPU Retrieval Transforms/filters data

Storage Storage
“Low-Level Retrieval”
Resources (Memory, CPU) Decodes stored data
Write Read
Many concurrent queries

I/O Performance
Block Count & Size History Blocks
Distributed Times Within Block (S-F, inserts)
Historian Retrieval Bottlenecks: Auto-Summary
Historian
Client
SQL Server
Throttles results sets > ~10K rows SQL
Insight
Server
May reduce rows with “WideHistory”
OLEDB REST SDK

Retrieval (Auto-Summary) “High-Level Retrieval”

Even faster with independent I/O channels Retrieval Transforms/filters data

Storage Storage Storage

“Low-Level Retrieval”
Resources (Memory, CPU) Decodes stored data
Write Read Write Read
Many concurrent queries

I/O Performance
Block Count & Size History Blocks
Distributed Times Within Block (S-F, inserts) Auto-
Summary
Standard
Historian Test: Average Flow Rate for a Month
select StartDateTime=MIN(StartDateTime), Historian
Client
TagName, Average=AVG(Value)
from History SQL
1M Insight
where Tagname like 'Flow1.PV' Server
and wwRetrievalMode='delta'
and DateTime>='2019-07-01' 1M
OLEDB REST SDK

and DateTime<='2019-08-01'
“High-Level Retrieval”
group by TagName 1M
Retrieval Transforms/filters data

Storage Storage
What is wrong with the query (not syntax)? 1M “Low-Level Retrieval”
Decodes stored data
Write Read Write Read

1. Load on many subsystems (slow)

2. Will not leverage auto-summary (slower)
1M History Blocks
3. Not time-weighted (incorrect in most cases) Auto-
Summary
Standard
Historian Test: Average Flow Rate for a Month
Historian
What is the optimal query? Client

SQL
1 Insight
select StartDateTime, TagName, Average Server

from AnalogSummaryHistory 1
OLEDB REST SDK
where Tagname like 'Flow1.PV'
and wwRetrievalMode='cyclic' 1.5K
Retrieval “High-Level Retrieval”
Transforms/filters data
and wwCycleCount=1
Storage Storage
and StartDateTime>='2019-07-01' 1.5K “Low-Level Retrieval”
and EndDateTime<='2019-08-01' Write Read Write Read Decodes stored data

Query Comparison 1.5K

1M History Blocks
Original: ~56 secs. New Query: <1 sec. Auto-
Summary
Standard
Cybersecurity Threat

• Recent Cybersecurity Survey

• Focused on operational technology (OT)
• At least one participant is a Wonderware Historian
customer (Taiwan Semiconductor)
• 701 total participants
• Highlights:
• 50% experienced plant downtime in the last 24 months
• 23% experienced a nation-state attack

© 2019 AVEVA Group plc and its subsidiaries. All rights reserved.

Approach to Security
A Holistic Approach Is Much More Than Technology

• Do you use encryption?

• Can we use Windows Authentication?
• Do we have to open firewall ports?
People SOP’s and
Policies Technology
Training Tools
Approach To Security

• Certified 1 October 2016

• First in industrial software
 Communications Security
Historian Insight
Client Browser
Always secure
by default
SQL Server HTTPS/TLS
HTTP*

Historian AVEVA Insight

(“tier 2”)

Replication (HCAL)*
HTTPS/TLS Secure by default in
Historian (“tier 1”)
WSP 2017 Update 3
HCAL* SuiteLink

Application Server InTouch

* advanced configuration required before 2017 U3
“[Insight] has already paid for itself...
I’m a [stubborn animal] for not doing this sooner.”
-- Insight Customer (in the first week)
 Online Maturity

2010 2011 2012 2013 2014 2015 2016 2017 2018

SmartGlance Historian Wonderware Newsfeed Hybrid Utilization Visualization

Online Online
(VM)

First managed solution First managedFirst

Search-based
Historian
nativeUnsupervised
cloud
selection
Historian
Hybrid
machine
on-premise/online
learning Integrated
Historian
utilization
Maps, Graphics
Geolocation
Significant New Insight Functionality
Machine Learning
+
• Supervised
• A human expert guides/configures
• Higher-value results
• Example: PRiSM
• Engineer identifies evaporator as
high-value equipment
• Finds related instruments
• PRiSM reports anomalies
• Unsupervised
• Automatically configured
• May report insignificant discoveries
• Example: Newsfeed
• No configuration required
• Reports anomalies
• Feedback part of training
Anomaly Detection
Integrated Publishing

Publish tag data as Insight data source

U3
Updated Insight Publisher
Insight Publisher: Look & Workflows
Search Changes
New Asset Page

Custom Graphic Auto-generated

Q4
 Insight Mobile

Dashboards Graphics

Q4 Q1
Winter 2019-20
Getting Started

Experience most end-user

functionality in a pre-configured
solution with live data

Create a free 45-day trial solution for your

own data. You must publish from your live
source before it is useful

© 2018 AVEVA Solutions Limited and its subsidiaries. All rights reserved.
Insight Data Source Changes

• Previously
• “Type” with “CSV/JSON”
• Could not mix “Publisher” & “CSV/JSON”
(normally)
• Processing time: up to 2 minutes
• Now
• Source of data no longer significant—mix & match
• Fixed property metadata (Description, Units, etc.) • Not a limitation for “Extended
must be “owned” by Publisher Properties” (Location, Alias, etc.)
• Processing time: instant
• Auto scale and 60% costs
Streamlined Re-Publish Workflow

• Skip the login, data source

selection steps
• Applies for adding/removing/
updating tags
• Included in WSP 2017
Update 3 SP1
• Available for download for
older InTouch
Login
Securing Internet Access
Firewall
• Primarily for blocking outside
access
• Normally restricted based on:
• Protocol (TCP, UDP)
• Source/destination IP address
• Port
Proxy
• Inspects request contents
• Normally restricted based on:
• Domain name
• Content (e.g. Java, ActiveX, Flash)
Insight Publishing With DMZ
Firewall

• Single outbound-only port (443)

•
Proxy
No inbound connection
• All traffic encrypted (TLS) Business

Firewall
• Anonymous HTTP proxy DMZ DMZ Secure Link
• Across a DMZ Firewall

• Works with any Publisher with proxy support Historian or

Insight Publisher
(Historian 2014 R2 SP1 Patch 02+)
• Destination must be AVEVA Insight Control
Limited Internet Access

© 2018 AVEVA Solutions Limited and its subsidiaries. All rights reserved.
Crossing The DMZ
Technical Details

• Captive, Application-specific Web Proxy

• Can’t be used to access other sites
• Whitelist
• Managed (managed by Insight)
• Signed (can’t be user-edited)
• Can be chained together
• Deployment
• Downloaded from Insight
• When connections fail, Publisher recommends it
Roadmap
2019 2020 2021
Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1

Historian 2017 U3 SP1 Historian 2020 Historian ?

• N-tier Replication • AVEVA Flex • Calculations (Unit Conversion)
• Citect Historian Migration • Time-sliced charts • Insight alarm view
On-Premise

• External alarm overlays

• InTouch alarm history
• Expanded Excel add-in

Insight
• Advanced alerts • Asset-based views • Alarm view • Production Day • Calculation • Recommended
• Single Sign-On • Asset hierarchy • Long-term • Calculations libraries actions
Managed

• Easier Power BI • Journal Performance • State-based News • Extensible News • Case detection
• News feedback news
• Mobile dashboards
• News alerts • Extension billing
• Mobile Graphics
• Time-sliced charts
Conclusions

• Our Vision & Strategy

• Has been consistent
• Translating into tangible value for Customers
• How can you better leverage it (Insight, “tiered”, Online)?
• Online
• Proven on-premises, online & hybrid platform
• Hybrid: for business user access, supply chain partners
• Pure: for stranded operational data
 linkedin.com/company/aveva
@avevagroup

ABOUT AVEVA
AVEVA is a global leader in engineering and industrial software driving digital transformation across
the entire asset and operational life cycle of capital -intensive industries.

The company’s engineering, planning and operations, asset performance, and monitoring and
control solutions deliver proven results to over 16,000 customers across the globe. Its customers are
supported by the largest industrial software ecosystem, including 4,200 partners and 5,700 certified
developers. AVEVA is headquartered in Cambridge, UK, with over 4,400 employees at 80 locations in
over 40 countries.
aveva.com

© 2019 AVEVA Group plc and its subsidiaries. All rights reserved.