1. Kent Weare
Integration MVP
Real World Industrial IoT

2. Agenda Intro to Industrial IoT Our Journey Our Reference Architecture
Demos

3. About Me Senior Enterprise Architect Azure (Integration) MVP Author
Multi-National Power Generation Company – Calgary, Alberta, Canada
Azure (Integration) MVP
Author
InfoQ – Cloud Editor
Blogger

4. What this talk is not about!

5. What this talk is about
Image Source:

6. What is Industrial IoT (IIoT)?
The Industrial Internet of Things (Industrial IoT) is made up of a multitude of devices connected by communications software. The resulting systems, and even the individual devices that comprise it, can monitor, collect, exchange, analyze, and instantly act on information to intelligently change their behavior or their environment -- all without human intervention.
Source:
Each windmill in the farm is equipped with many sensors and software that allows it to communicate with all the other windmills and even request repairs. To optimize power generation, each windmill automatically adjusts its settings and behavior based on data it receives and processes from the system, such as changes in wind speed and direction. Such environmental changes may result not only directly from the weather, but also from the behavior of other windmills.

7. IIoT is not a new discipline…
Each windmill in the farm is equipped with many sensors and software that allows it to communicate with all the other windmills and even request repairs. To optimize power generation, each windmill automatically adjusts its settings and behavior based on data it receives and processes from the system, such as changes in wind speed and direction. Such environmental changes may result not only directly from the weather, but also from the behavior of other windmills.

8. It is the evolution of existing technologies…

9. IIoT’s roots are in SCADA…
What is SCADA?
SCADA is an acronym for Supervisory Control and Data Acquisition. It is software and hardware that can be used for process control and gathering real time data from remote locations for exercising control on equipment and conditions.
Orlando Moreno

10. SCADA System
Image Source:

11. Are IIoT and SCADA the same?
Image Source:

12. No, their prescribed implementations are very differentvs
Image Source:
But they do address some of the same use case.

13. SCADA Systems are highly dependent on:
Provisioning of Infrastructure
Multi-level Networks
Firewalls
Expensive, proprietary software or protocols
An unwillingness to change
Which ultimately leads to…

15. Image Source: http://ww3. hdnux. com/photos/06/76/34/1828634/7/920x920

16. Ask BMW about their connected car…
Am I being too harsh?
Ask BMW about their connected car…

17. Our Journey…
Image Source:

18. What prompted our change in Architecture?
Image Source:

19. Image Source: https://website-designs

20. We could have delivered a point solution…

21. But opted for a flexible, layered architecture…
Delivered
Data Collection
Real Time Analytics
Future Opportunity
Pro-active Alerts
Pre-emptive Maintenance
Work Order Integration

22. Disaster Recovery

23. Disaster Recovery

24. But can traditional SCADA and IOT Architectures co-exist?
Arsenal-Tottenham 

25. But can these Architectures co-exist?
It would not be feasible, from a time or cost perspective, to retro fit everything. Instead we replace the components that make sense within the context of time and budget

26. What did we implement?

27. A Hybrid messaging platform that:
Feeds existing Historian and allows Plant Operation personnel to monitor that state of devices and Power Generation output (MW)
Provides telemetry for downstream reporting, compliance and commercial purposes.
Real-time visibility for Field IT Support
Flexibility to evolve

28. What were our timelines?

29. January – POC, Architecture Reviews, Contract Negotiations
February/March – Build and Testing
April – Parallel Systems Cutover
> 56 million Device Reads processed
May – Decommission Legacy Solution

30. Our Reference Architecture

31. Microsoft Azure IoT Services
Devices
Device Connectivity
Storage
Analytics
Presentation & Action
Event Hubs
SQL Database
Machine Learning
App Service
Service Bus
Table/Blob Storage
Stream Analytics
Power BI
External Data Sources
DocumentDB
HDInsight
Notification Hubs
Data Factory
Mobile Services
BizTalk
{ }

32. Event Hubs Cloud-scale telemetry ingestion
Build 2014
4/11/2018
Event Hubs
Cloud-scale telemetry ingestion
Compatible with more than a million publishers supporting HTTP, AMQP and MQTT
Ingress millions of events per second
SAS based security, with unique token per publisher
Configurable data retention (1-30 days)
Low latency (<10 ms for volatile data)
Pluggable with other cloud services like Stream Analytics
Event Hub
© 2014 Microsoft Corporation. All rights reserved. Microsoft, Windows, and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

33. Stream Analytics Real-time analytics for Internet of Things solutions
Build 2014
4/11/2018
Stream Analytics
Stream Analytics
Real-time analytics for Internet of Things solutions
Calculation Engine
Sum, Average, Count, Max, Min
Intelligent Windows
Tumbling
Hopping
Sliding
Stream millions of events per second
Mission critical reliability, performance and predictable results
Rapid development with familiar SQL-based language
Routes information
© 2014 Microsoft Corporation. All rights reserved. Microsoft, Windows, and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

34. Stream Analytics - Windows
Build 2014
4/11/2018
Stream Analytics - Windows
Tumbling Window
Hopping Window
Sliding Window
SELECT sensorId, COUNT(*) AS Count FROM SensorReadings TIMESTAMP BY time GROUP BY sensorId, TumblingWindow(second, 10)
SELECT sensorId, MIN(temp) as temp FROM SensorReadings TIMESTAMP BY time GROUP BY sensorId, SlidingWindow(second, 5) HAVING MIN(temp) > 75 
SELECT sensorId, COUNT(*) AS Count,
AVG(temp)
FROM SensorReadings TIMESTAMP BY time
GROUP BY sensorId,
HoppingWindow(second, 10 , 5)
Source:
Windowing
Windowing is a core requirement for stream processing applications to perform set-based operations like counts or aggregations over events that arrive within a specified period of time. ASA supports three types of windows: Tumbling, Hopping, and Sliding.
Tumbling Windows define a repeating, non-overlapping window of time.  
Like Tumbling Windows, Hopping Windows move forward in time by a fixed period but they can overlap with one another.
With a Sliding Window, the system is asked to logically consider all possible windows of a given length and output events for cases when the content of the window actually changes – that is, when an event entered or existed the window.
© 2014 Microsoft Corporation. All rights reserved. Microsoft, Windows, and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

35. Service Bus Topics Pub-Sub Messaging Load-levelling
Build 2014
4/11/2018
Service Bus Topics
Pub-Sub Messaging
Load-levelling
Output for Stream Analytics
Cost effective Cloud Messaging
© 2014 Microsoft Corporation. All rights reserved. Microsoft, Windows, and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

36. Power Bi Datasets Reports Dashboards
Real Time & Scheduled Update
Self-service data exploration and analysis
Office-like experience
Cloud and Mobile friendly

37. BizTalk Server On-Premises Integration Engine Protocol Mediation
Data format transformation
Routing

38. Azure Logic Apps Cloud-based Orchestration Notifications (SMS/Email)
SaaS Connectivity

39. Demo 1 - Telemetry
Image Source:

40. Demo 1 – Telemetry Architecture
Our core use case – deliver telemetry

41. Demo 1 – Telemetry Architecture
C# Simulator – Publish to Event Hub
Create Event Hub Connection
Instantiate Message
Stream Analytics
Populate Randomized Tag Values
Serialize and Send (Async)

42. Demo 1 – Telemetry Architecture
Message Payload
Stream Analytics

43. Demo 1 – Telemetry Architecture
Stream Analytics Input
Stream Analytics

44. Demo 1 – Telemetry Architecture
Stream Analytics Output
Stream Analytics

45. Demo 1 – Telemetry Architecture
Stream Analytics Query
Stream Analytics

46. Demo 1 – Telemetry Architecture
BizTalk Processing
Receive Port
Send Port
BizTalk
Pipeline
(XML-CSV)
Service Bus Topic
BizTalk
Pipeline
(JSON-XML)
BizTalk
MsgBox
Historian
File Collector
Stream Analytics

47. Demo 1 – Telemetry Architecture
BizTalk Output (de-identified)
Stream Analytics

48. Demo 2 – Real Time Visualizations
Image Source:

49. Demo 2 – Real Time Visualizations Architecture

50. Demo 2 – Real Time Visualizations
Stream Analytics Inputs
Stream Analytics

51. Demo 2 – Real Time Visualizations
Stream Analytics Inputs – Reference Data
Stream Analytics

52. Demo 2 – Real Time Visualizations
Stream Analytics Outputs
Stream Analytics

53. Demo 2 – Real Time Visualizations
Stream Analytics Output – Power BI
Stream Analytics

54. Demo 2 – Real Time Visualizations
Stream Analytics Query
Stream Analytics

55. Demo 2 – Real Time Visualizations
Power BI
Stream Analytics

56. Demo 3 – Detecting Absence of Event(s)
Stream Analytics

57. Demo 3 –Absence of Event(s) Architecture
Stream Analytics

58. Demo 3 –Absence of Event(s) Architecture
Stream Analytics Query
Stream Analytics

59. Demo 3 –Absence of Event(s) Architecture
Logic Apps
Stream Analytics

60. Demo 3 –Absence of Event(s) Architecture
Azure Functions
Stream Analytics

61. Demo 3 –Absence of Event(s) Architecture
Text Message
Stream Analytics

62. Monitoring
Stream Analytics

63. Monitoring with BizTalk360
Web Endpoint Monitoring
BizTalk360
Process Monitoring
BizTalk Server
BizTalk360

64. Monitoring with BizTalk360
BizTalk360 Web Endpoint Monitoring
Stream Analytics

65. Monitoring with BizTalk360
API Management Policy
Stream Analytics

66. Monitoring with BizTalk360
Stream Analytics Query
Stream Analytics

67. Monitoring with BizTalk360
BizTalk Receive Location
Stream Analytics

68. Monitoring with BizTalk360
BizTalk360 Process Monitoring
Stream Analytics

70. Lessons Learned Drive adoption through POCs (Fail Fast/Test & Learn)
Decision makers like the option to say ‘No’ – give them that opportunity
Frame benefits holistically
Involve Security early
Write ASA queries defensively to avoid job failures
Don’t tie Power BI accounts to a user
Decouple data collection and analytic streams
Leverage Event Hub ‘re-play’ capabilities
Power BI can only be configured after you have data
Watch for ACS (BizTalk 2013)
Don’t forget about monitoring

71. Image Source: https://memoirsofanadventurer

72. What’s Next? Challenge Status Quo Focus on plant modernization
Need flexibility (investments/divestures)
Reduce dependency on expensive proprietary software
OPC UA – AMQP Bridge
Build Historian “like” alternative
Build out advanced analytics capabilities

73. Questions? Blog: http://www.MiddlewareInTheCloud.com Twitter: @wearsy
InfoQ:
BMW Connected Car: (36m:25s mark)