users conference 2016 - cdn.osisoft.com
TRANSCRIPT
![Page 1: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/1.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
![Page 2: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/2.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Presented by
Get More From Your
PI System Data With
Advanced Analytics
Iain Allen, Senior Manager, Mining IT, Barrick Gold
Sameer Kalwani, Co-Founder & VP of Product, Element Analytics
![Page 3: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/3.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
The Gold Mining Business
3
Tough and getting tougher
Cost control is paramount
Less and less bang
for the Exploration buck!
![Page 4: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/4.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
#1
Barrick in the Gold Mining Business
4
#1Company
Market Cap
(US$Bn)
2015 2014
Barrick 15.7 7.4
Newmont 14 8.57
Goldcorp 13.19 10.99
Newcrest 9.93 6.44
Polyus 8.79 8.64
Agnico 7.93 5.45
Anglogold Ashanti 5.58 3.48
Gold Fields 3.25 2.13
Yamana 2.91 1.63
Eldorado 2.25 2.21
#1
![Page 5: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/5.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
How Did Barrick Become #1 Again?
5
2012• 25 operating mines
• ~7.5 million ounces produced
• AISC US$915
• Reserves 104.1 million ounces
2015• 12 operating mines
• ~6.25 million ounces produced
• AISC US$831
• Reserves 93 million ounces
![Page 6: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/6.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
All In Sustaining Costs (AISC)
6
• Does not include
• Capex for new mine
construction
• Debt repayment
![Page 7: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/7.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Barrick’s Goal – Gold Price Agnostic
“Our aspiration is to achieve all-in sustaining costs below $700 per ounce by 2019.”
The Law of Diminishing Returns
2013
Actual $915
2014
Guidance 920-980
Actual $864
2015
Guidance 860-895
Actual $831
830-860
![Page 8: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/8.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016 8
“It is crucial to the future of the company that
Barrick become a Data-Driven Decision-making
Organization”
Jim Gowans, Chief Operating Officer, 2014
![Page 9: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/9.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
The Digitization of Barrick
9
Sub $800
AISC
Best in Class
Dynamic Mass-
Energy BalanceBusiness
Enablement and
Simplification
Production
Costing Water
ManagementCondition-Based
Maintenance (CBM)
![Page 10: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/10.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016 10
“Our aspiration is to achieve all-in sustaining
costs below $700 per ounce by 2019.”
![Page 11: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/11.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Pueblo Viejo Mine
• Barrick’s newest mine
• Produces 800k ounces of
gold per year
• Most advanced PI site
• Doing very innovative work
with PI on Energy and CBM
• Have a positive attitude
toward new ideas
11
![Page 12: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/12.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Producing Analytics
12
Descriptive Analytics Predictive Analytics
7, 11%
7, 10%
41, 61%
12, 18%
![Page 13: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/13.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Demo: Descriptive
Analytics
13
![Page 14: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/14.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
How we got to Descriptive
Analytics Quickly
14
![Page 15: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/15.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Emerging Categories Create a lot of Noise
15
Internet of Things
Big Data
Industrial Internet
Industry 4.0IT/OT Convergence
![Page 16: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/16.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Looking for Solutions That Make
Our People and Equipment More Effective
16
Operational
Information• Maintenance
• Operator Action
• Machine Data
Business Data• Market Pricing
• Supply Chain• Financial Optimization
IT Capabilities• Big Data
• Internet of Things
• Analytics
![Page 17: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/17.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Application
Connection
Predictive Modeling
OSIsoft PI System
(Analytics)
Element Analytics Platform Architecture (an OSIsoft ConnectedApp)
17
OSIsoft PI System
(Operational)On
Pre
mis
e
(Additional Data Historians)
Clo
ud
External Data Sources
(e.g., LIMS, EAM, CMMS, ERP)
Data Preparation
Data Standardization
EngineContextualization Engine Prepped Data Subsets
Power BI
Azure ML
PI Coresight
Predictive Web
Services
Predictive
ModelsData Trust Assurance
Element Analytics Installed on Customer Tenant on Microsoft Azure
Event Hub
PI
Integrator
![Page 18: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/18.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Demo: Element Platform
18
![Page 19: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/19.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Contextualization – Easily Surface Events, and Label them as Event Frames.
We Use All Assets of an Asset Template + All Historical Data to Surface Events
19
![Page 20: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/20.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
PI Integrator
20
![Page 21: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/21.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Getting Predictive Results
21
![Page 22: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/22.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Every Minute Counts!
Now That We Can Measure It And See It, We Can Do Something About It
Identifying the Opportunity
• With PowerBI, we identified
downtime issues with the
limestone crushing system.
• The biggest 3 systems on that
circuit are:
– Limestone crushers
– Lubrication system
– Conveyor
How to Solve the Problem
• Requirement:
– need at least 1 week notification in
advance to predict potential faults
• Challenge:
– Scheduling more frequent
maintenance doesn’t scale.
• Solution:
– Need to catch faults before they
happen
– Need Data Science to create
predictive models
22
![Page 23: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/23.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Approach to build a model to predict unplanned failures
in limestone circuit a week in advance
23
Prepared
Data
(PI + External Data)
Generated Features Algorithms
Predictive Model
Streaming
PI Data
Predictive
Web ServiceReports
![Page 24: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/24.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Identifying Features
• Dynamic Features
- Traditional tags that change based off the conditions of
the system it is monitoring
• Alarm Features
- Binary status/state tags that indicate on/off conditions
of the system it is monitoring
![Page 25: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/25.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Dynamic Features
Crusher
• Current
• Main Drive Coupling Zero
Speed
• Power Used
• Area Receiver Pressure
• Pulse Air Receiver
Pressure
• MPS Pressure
25
Conveyor
• Head Temperature
• Weight Scale
• Conveyor Current
Lubrication System
• Discharge Pressure
• Backup Pump Pressure
• Oil Res Temperature
• Oil Rtn Line Temperature
• Primary Pump Pressure
• Centrifugal Filter Pump
• Cooling Fan
• Pump
Std Dev. Min Max … Mean
![Page 26: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/26.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Alarm Features
Crusher
• Blower Filter Warning
• UPS Alarm
• Dirty Filter Warning
• Plugged Filter Warning
• Screen Diverter Feed
Chute Level
26
Conveyor
• Conveyor RIP Detect
• Conveyor Tail Speed
• Conveyor Misalign
• Conveyor Pullcord
Lubrication System
• Oil filter Valves Closed
• High Oil Temp Alarm
• Low Oil Temp Alarm
• Low Oil Level Alarm
• Oil Filter Valve Open
• Filter Plugged
Daily Counts
![Page 27: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/27.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Splitting the Data to Build the Model
27
PI + RtDuet Data
Training Data(also used for cross validation)
Test Data
1/1/2015 11/30/2015
1/1/2015 11/30/20159/10/2015
![Page 28: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/28.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Decision Trees
Automatically identify
ways to divide data.
![Page 29: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/29.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Support Vector Machines
Finds Non-linear
separation in data
![Page 30: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/30.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
How to Measure Predictive Model Performance
The Ecosystem of a
Predictive Mode
• True Positive
• True Negative
• False Positive
• False Negative
• Precision
– TP/(TP+FP)
• Recall
– TP/(TP+FN)
30
Actual
Faults
Predicted
Faults
Predicted
Faults
Actual
Faults
Actual
Faults
Predicted
Faults
![Page 31: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/31.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Crusher Predictive Initial Results
31
TP - Correctly predicted faults FP - Incorrectly predicted faults
TN - Correctly predicted non-faults FN - Actual fault, but not predicted
77
Precision
7
12
Recall
7
7
41
12
Crusher System Predictive Results
Number of assets trained upon 1
Actual number of days that are predictive of
faults
19
Actual number of days that are predictive of
non-faults
48
Number of faults predicted by model 14
% predicted faults that actually occur 7 (50%)
% of predicted faults that are non-faults 7 (50%)
% of total faults that were predicted 7 (37%)
![Page 32: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/32.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Lube System Predictive Initial Results
32
8
11
35
10
Lube System Predictive Results
TP - Correctly predicted faults FP - Incorrectly predicted faults
TN - Correctly predicted non-faults FN - Actual fault, but not predicted
8
11
Precision
810
Recall
Number of assets trained upon 1
Actual number of days that are predictive of
faults
18
Actual number of days that are predictive of
non-faults
46
Number of faults predicted by model 19
% predicted faults that actually occur 8 (42%)
% of predicted faults that are non-faults 11 (58%)
% of total faults that were predicted 8 (44%)
![Page 33: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/33.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Conveyor Predictive Initial Results
33
TP - Correctly predicted faults FP - Incorrectly predicted faults
TN - Correctly predicted non-faults FN - Actual fault, but not predicted
26
15
13
11
Conveyor System Predictive Results
26
15
Precision
26
7
Recall
Number of assets trained upon 1
Actual number of days that are predictive of
faults
33
Actual number of days that are predictive of
non-faults
28
Number of faults predicted by model 41
% predicted faults that actually occur 23 (67%)
% of predicted faults that are non-faults 17 (33%)
% of total faults that were predicted 26 (79%)
![Page 34: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/34.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Recap
34
![Page 35: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/35.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Recap: The Analytics Journey Requires the Right Data to Enable it
35
Basic Analytics
Descriptive
Analytics
• Tag-by-tag querying
Diagnostic
Analytics
Predictive
Analytics
Prescriptive
Analytics
Today’s Data• Connected to multiple systems
• Site to site data variance
• Varying naming conventions
• Inconsistent units of
measurement
Analytics Maturity
• What happened?
• Provides info about
failures and other past
problems
Getting valuable foresight requires
increasingly proactive, advanced
analytics, supported by the right kind
of data.
• When and Why did it
happen?
• Provides insight and
visibility into what can be
improved where
Descriptive
Data• Standardized PI data
in an organized,
consistent form
• SME knowledge
Asset Framework
Diagnostic Data• Contextualized data
with labeled events
• Failure data,
maintenance data,
weather data,
geological data
Event Frames
Predictive Data• Trusted data (uniform
sensor coverage, no
noise, calibration, drift
issues)
• SME knowledge,
engineering data,
forecast data
Sensor Audit
• What will happen?
• Provides predictions that
lower maintenance costs,
optimize efficiency, and
improve worker safety
• How can we make it
happen?
• Provides recommendations
for the best course of action
to achieve desired
outcomes
Prescriptive
Data• Join multiple
operational data
sources (e.g.
Maintenance activity)
• Additional workflow
data, such as
maintenance
personnel data, supply
chain data, inventory
data
Integrations/Connections
![Page 36: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/36.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Lessons Learned
• Involve the site team from the beginning
• Provide as much PI data as possible
• Provide detailed descriptions of your PI tags
• Provide any supporting data – RtDuet, Ivara
• Have a follow-up workshop to present and discuss
the results
36
![Page 37: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/37.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Contact Information
Iain Allen
Senior Manager, Mining IT
Barrick Gold
37
Sameer Kalwani
Co-Founder & VP of Product
Element Analytics
37
![Page 38: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/38.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Questions
Please wait for the
microphone before asking
your questions
Please don’t forget to…
Complete the Online Survey
for this session
State your
name & company
38
![Page 39: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/39.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Thank You
Talk to us about:
Predictive Analytics for Industry - http://bit.ly/1UfGnPt
The Future of Industrial Big Data and Its Data Tech Stack - http://bit.ly/1mg2CsY
![Page 40: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/40.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
![Page 41: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/41.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
![Page 42: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/42.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
![Page 43: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/43.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Getting Ready for Analytics:
Moved PI Tags into Consistent Asset Templates
43
We were able to get 7700 tags across 11 asset templates completed in 3 hours,
using the Element Analytics Platform
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
![Page 44: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/44.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016 44
Created Asset Hierarchies
![Page 45: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/45.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Questions
Please wait for the
microphone before asking
your questions
Please remember to…
Complete the Online Survey
for this session
State your
name & company
45
http://ddut.ch/osisoft
search OSISOFT in the app store
![Page 46: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/46.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016
Thank You
![Page 47: USERS CONFERENCE 2016 - cdn.osisoft.com](https://reader034.vdocuments.mx/reader034/viewer/2022051510/627e92a5cfe8af464956a73e/html5/thumbnails/47.jpg)
© Copyright 2016 OSIsoft, LLCUSERS CONFERENCE 2016