This document discusses how smart manufacturing and artificial intelligence of things (AIoT) can help drive digital transformation. It provides examples of how IoT solutions have helped various companies reduce costs and improve operations. It then discusses key concepts in smart manufacturing like the intelligent edge, cloud computing, and different waves of innovation with IoT, edge, and AI. The document outlines Microsoft's IoT portfolio and reference architecture for smart manufacturing. It also describes various Azure IoT capabilities and solutions like IoT Hub, IoT Edge, Time Series Insights, and preconfigured solutions for predictive maintenance, remote monitoring and connected factories. Finally, it discusses how machine learning can address supply chain optimization, predictive maintenance, anomaly detection, production scheduling and demand
5. Open source support beyond your imagination
Applications
Infrastructure
Management
Databases and
middleware
App frameworks
and tools
DevOps
6.
7. Connected chillers are
back online 9x faster than
unconnected equipment,
avoiding more than
$300,000 in hourly
downtime costs
Data from sensors and
systems to create
valuable business
intelligence and reduce
downtime by 50%
Reduced its accident rate
by 25% and fuel usage by
20%, reporting annual
savings of $1.8 million
Cut down-time cut for
each packaging line by up
to 48 hours, saving
€30,000 for customers
Keeping farmers informed
about irrigation, disease
control diseases, and pest
has led to increased yields
of 30%, and a 20%
reduction in water use
Rolls Royce “power by the
hour” model provides
maximize availability by
cutting fuel consumption by
1% and up to $250,000 per
plane, per year.
Access to production and
supply chain data worldwide,
reduced downtime costs by
as much as $300,000 per day
Licorice extruders on
Twizzler’s production line
are performing at peak
optimization, saving
over $500K/year on
materials alone
Enabled customers to
transport more than 1M
additional tons of
cargo, and reduce fuel
consumption by 17%
8.
9. Cloud
Globally available, unlimited compute resources
Waves of Innovation
IoT
Harnessing signals from sensors and devices,
managed centrally by the cloud
Edge
Intelligence offloaded from the cloud to IoT
devices
AI
Breakthrough intelligence capabilities, in the cloud
and at the edge
12. Intelligent Edge
High-speed data processing,
analytics and shorter response
times are more essential than ever.
Intelligent Cloud
• Business agility and scalability: unlimited
computing power available on demand.
Intelligent Edge
• Can handle priority-one tasks locally
even without cloud connection.
• Can handle generated data that is too
large to pull rapidly from the cloud.
• Enables real-time processing through
intelligence in or near to local devices.
• Flexibility to accommodate data privacy related
requirements.
17. Digital transformation is real – and it’s happening now
83%
of manufacturers said
that selling products
as services increases
profits
72%
of field service
organizations treat
service as a profit
center
Create service-
based business
models
40%
of industrial
manufacturers use
digital technologies to
monitor products sold
to customers
73%
of manufacturing
executive are launching
IoT initiative in 20171
Enhance the
customer
experience
80%
of manufacturers
expect that improved
factory connectivity
will help them increase
output
levels
35%
of manufacturers
currently collect and use
data generated by smart
sensors to enhance
manufacturing
Innovate
faster
1. Source: State of the Market: Internet of Things 2017, Verizon
18. CRM
SRM
ERP
…
Process
interoperability (IT)Information
Flow (OT)
Collaborative
equipment Ecosystem
Interaction
Mixed/Augmented
Reality
Simulation
Industrial IoT
Cloud
3D Printing
Big Data
and
Analytics
Vertical integration | Horizontal integration | End-to-end engineering
Industry 4.0 Design Principles
Industry 4.0
Machine
Learning
Autonomous Robots
Cybersecurity
BlockChain
19. IoT tangible results … moving to SMB soon
Gathers data from sensors and
systems to create valuable
business intelligence and
reduce downtime by 50%
Cutting fuel usage by 1%
could save $250,000 per
plane per year
Chillers now run 9x faster
than unconnected equipment,
avoiding more than $300,000
in hourly downtime costs
Improves access to production
and supply chain data
worldwide, reducing downtime
costs by as much as $300,000
per day
20. Connecting and controlling devices
Device Registry/Management
Collecting and managing data
Stream
Analytics
External
Data Sources
Transforming real-time data to business insights
Gateway
Delivering insights to decision makers Performing advanced analytics
Data Mash Ups
Data Factory
IoT agent
Event Hubs & Service Bus
IoT Hub
Data Storage
SQL DB Table / Blob Storage
Business Intelligence
Power BI
Azure Websites
Mobile Services
Predictive Maintenance Event
Analysis
Generate
Prediction Models
Prediction
Models
Training
Models
Machine Learning
(Azure ML)
HDInsight
(Hadoop)
ML Studio
External
Services
Local Technician
Remote Expert
Azure
Service Bus
HoloLens
OS
LOB Systems of Record
MESERP
21. Service Ticket
History (CRM)
Other CAD
Systems
Long
Distance
Engine
? More
Aluminum?
Carbon
Fiber
Parts
?
Placeholder Placeholder Placeholder
Connected Product Innovation
Short Range Engine Short Range Training
Understand exactly how products are
performing in the field
1
FLIGHT DATA
Investigate issues and redesign as necessary
4
Anticipate new use cases and optimize product performance
5
Change
Requests
IoT
Trends
Service Ticket
History (CRM)
010101001001010101001001
010100100101010100100100DIGITALSTREAM
Engineer /
Design Change
010101001001010
010100100101010DIGITAL STREAM
Idea Management
Center
FACTORIES PRODUCING FUEL PUMP FACTORIES PRODUCING FUEL PUMP
USA GERMANY
Engineering RequestsSupplier Investigations
Vendor
Change Request
Sales/Service
Campaigns
Combine product data across all relevant sources
2
Run simulations to predict performance issues
and identify opportunities for improvement
3
22. Service Ticket
History (CRM)
Other CAD
Systems
Long
Distance
Engine
? More
Aluminum?
Carbon
Fiber
Parts
?
Placeholder Placeholder Placeholder
Connected Product Innovation
Short Range Engine Short Range Training
Understand exactly how products are performing
in the field
• Know how frequently the product is being used
• Find usage and failure rates for specific parts or
features
• Discover 360°view of product performance
compared to design
1
FLIGHT DATA
Investigate issues and redesign as necessary
• Resolve instances of over- and under-engineering to optimize product
performance/cost ratio
• Design different product configurations to better fit usage patterns
• Support redesign of areas with regular failures
• Manage vendor lists and track standard performance across suppliers
4Anticipate new use cases and optimize product performance
• Save time and money by combining preventative maintenance with replacement
of redesigned parts
• Refine products to fit stakeholder needs at product launch and throughout the
product lifecycle
• Design new products lines and offer new services, like training, based on product
insights
5
IoT
Trends
Service Ticket
History (CRM)
Power BI Embedded within
ISV XXX
Cortana Intelligence Suite
Change
Requests
Develop new streams of revenue, like new product models, lines and services
Coordinate one-off changes and preventative repairs with existing maintenance
schedules
Give technicians detailed repair walk-throughs using wearables and augmented
reality
Improve collaboration and design with wearables and augmented reality
Streamline cross-org communication and automate regular handoffs
between departments
Visualize data from PLM and CAD systems in a unified view
CAD on Azure
Big Compute
Connected Product Innovation
Connected Field Service
Hololens
Azure IoT
Acceleration
Solutions
010101001001010101001001
010100100101010100100100DIGITALSTREAM
Vendor
Change Request
Engineer /
Design Change
010101001001010
010100100101010DIGITAL STREAM
Sales/Service
Campaigns
Idea Management
Center
FACTORIES PRODUCING FUEL PUMP FACTORIES PRODUCING FUEL PUMP
USA GERMANY
Power BI Cortana Intelligence Suite
Engineering RequestsSupplier Investigations
Hololens
Combine product data across all relevant sources
• Aggregate data from silos such as maintenance records, design
specifications, and diagnostics
• Analyze data across vendors, part specifications, and other external
sources
• Integrate stakeholder feedback on existing and suggested features
for 360°view of customer
2
Predict disruptive events and identify opportunities
• Collaborate across teams and geographies with product-
focused tools
• Determine if products meet design benchmarks
• Compare quality of variants using data-driven analysis to
guide design changes
3
Product Insights
Azure Data Lake
Connected Product Innovation
23. IoT projects are complex
New skill sets
Time-consuming setup and
integration
Heavy up-front investment
24. Smart Manufacturing Reference Architecture
Field
Gateway
IoT Hub
Data Ingestion
Device Registry
Device Management
HTTP/s, AMQP, MQTT
Microsoft Azure
Office365NUI
PowerBI
Enterprise
Business
Systems
PLM ERP MES CRM
Stream Analytics
HD Insight
Machine Learning
SQLTable/Blob
Storage
{ }
DocumentDB Data Lake
ID
Devices
Sensors
Field
Gateway
OPC
Publisher
SCM
Event Hub
Data Catalog
Data Factory
Information Management
Big Data Storage
Azure Biztalk
Logic Apps
Analytics and Machine Learning
Business Integration
and Gateways
Visualization
25. User experiences
Dashboards Mixed Reality Interactive speech Gestures
Preconfigured Azure IoT Suite and SaaS applications
Remote monitoring Predictive maintenance Connected factory Microsoft IoT Central
Analytics and artificial intelligence
HDInsight Machine
Learning
Data Lake
Analytics
Azure Time
Series Insights
Bot
Framework
Operations
Technology (OT)
Enterprise business processes
PLM ERP SCM MESCRM
Home
EdgeAnalyticsDeviceAgent
StreamAnalytics
onedge
Logic
Apps
API
integration
BizTalk
Services
Azure StackSQL Server
Enterprise IT Business integration
Hot path analytics and application platform
Cold path analytics and storage
Stream Analytics Event Hubs Service Fabric (Actors) Functions
Data Factory DocumentDB SQL Database Data Lake Store
Field Gateway
Cloud Gateway
Azure IoT Gateway
Azure IoT
Hub
(device
provisioning)
Cognitive
Services
26. User experiences
Dashboards Mixed Reality Interactive speech Gestures
Preconfigured Azure IoT Suite and SaaS applications
Remote monitoring Predictive maintenance Connected factory Microsoft IoT Central1
Analytics and artificial intelligence
HDInsight Machine
Learning
Data Lake
Analytics
Azure Time
Series Insights
Bot
Framework
Operations
Technology (OT)
Enterprise business processes
PLM ERP SCM MESCRM
Home
Logic
Apps
API
integration
BizTalk
Services
Azure StackSQL Server
Enterprise IT Business integration
Hot path analytics and application platform
Cold path analytics and storage
Stream Analytics Event Hubs Service Fabric (Actors) Functions
Data Factory DocumentDB SQL Database Data Lake Store
Field Gateway
Cloud Gateway
Azure IoT Gateway
Azure IoT
Hub
(device
provisioning)
Cognitive
Services
StreamAnalytics
onedge
DeviceAgent
27. User experiences
Dashboards Mixed Reality Interactive speech Gestures
Preconfigured Azure IoT Suite and SaaS applications
Remote monitoring Predictive maintenance Connected factory Microsoft IoT Central1
Analytics and artificial intelligence
HDInsight Machine
Learning
Data Lake
Analytics
Azure Time
Series Insights
Bot
Framework
Operations
Technology (OT)
Enterprise business processes
PLM ERP SCM MESCRM
Home
Logic
Apps
API
integration
BizTalk
Services
Azure StackSQL Server
Enterprise IT Business integration
Hot path analytics and application platform
Cold path analytics and storage
Stream Analytics Event Hubs Service Fabric (Actors) Functions
Data Factory DocumentDB SQL Database Data Lake Store
Field Gateway
Cloud Gateway
Azure IoT Gateway
Azure IoT
Hub
(device
provisioning)
Cognitive
Services
StreamAnalytics
onedge
DeviceAgent
28. Microsoft’s IoT portfolio
Microsoft’s vision is to democratize IoT by allowing everyone to access
the benefits of IoT and provide the foundation for digital transformation
Fully managed SaaS
Best used when you need to get started quickly
with minimal IoT experience
Microsoft
IoT Central
Securely distribute
cloud intelligence
locally, quickly, and
at scale
Hybrid – Edge
Analytics
Azure IoT
Edge
+
Customizable PaaS
Best used when you need a lot of control over
your IoT solution
Azure IoT
Suite
29. Azure IoT Acceleration Solutions
Easily ingest data and create command & control responses
Analyze incoming data via stream processing and machine learning
Integrate with existing business systems and automate workflows
Leverage PowerBI for rich dashboards and visualization
Deploy quickly with preconfigured Solutions
Predictive Maintenance
Remote Monitoring v2
Securely connect to and manage devices and gateways
Connected Factory
30. A new preconfigured solution to connect your OPC UA and OPC Classic devices to Azure to:
• Monitor factory, production lines, station OEE, and KPI values
• Analyze the telemetry data generated from these devices using Azure Time Series Insights
• Act on alerts to fix issues
Connected Factory Solution
IoT Hub
VM
Linux VM (with multiple assembly lines)
Web App hosting
Solution Dashboard &
OPC UA Client
OPC UA Server
OPC UA Server
OPC UA Server
Gateway SDK with
OPC Proxy &
OPC Publisher
Modules
MES
Simulation
(OPC UA Client)
Telemetry path
Command &
Control path
Time Series Insights
31. Designed for IoT
Connect up to 10 million devices
Cloud-scale messaging
Device-to-cloud and Cloud-to-device
Durable messages (at least once semantics)
Per-device authentication
Individual device identities and credentials
Multi-protocol support
Natively supports AMQP, HTTP, MQTT
Designed for extensibility to custom protocols
Service assisted communications
Secure bi-directional communication
Command and control
Cloud-facing telemetry ingestion
Delivery receipts, expired messages
Device communication errors
Connection multiplexing
Single device-cloud connection for all
communications (C2D, D2C)
Multi-platform
Device SDKs available for multiple platforms
(e.g. RTOS, Linux, Windows)
Multi-platform Service SDK
32. Azure IoT Edge
• Low-latency (important in safety scenarios, such as stop the machine immediately)
• Intermittent connectivity (remote mining, oil, factories)
• Integrate sensor data from different devices
• Filter or aggregate telemetry data before sending it to the cloud
• Transform raw input from sensors to meaningful information
On Premise
Azure IoT Edge
Devices
RTOS,Linux,Windows,
Android,iOS
Stream
Analytics
Custom
OPC UA or
Modbus
ingestion
module
Azure
Functions
Machine
Learning
Cognitive
Services
Container
Management
Functions
Runtime
Device Twin
Local
Storage
IoT Hub
Device
Twin
Stream
Analytics
Machine
Learning
Power BI
35. Brownfield Enabled
<$500
+ No changes to
machines required!
Consistent,
compatible data
model for all
machines,
plus security!
No compatibility!
No winner in the field bus/industrial
ethernet wars!
36. OPC
Publisher
Device
Connectivity & Management
Analytics &
Operationalized Insights
Presentation &
Business Connectivity
IndustrialDevices
(OPCandOPCUAServers)
Hot Path Analytics
Azure Stream Analytics, Azure Storm
Presentation & Business
Connections
Websites, Mobile Services
Dynamics, BizTalk Services,
Notification Hubs
Cloud Gateway
IoT Hub
Field
Gateway
OPC API & Client
Hot Path Business Logic
Service Fabric & Actor Framework
Azure IoT Accelerator Solutions Supports OPC UA
37. Azure Time Series Insights (提供強大時序資料處理能力)Devices
Storage
• IoT SCALE TIME-SERIES DATA STORE Schema-less store, just send data
Easy IoT Hub connection
Store, query and visualize billions of events
Simple and fast navigation
38. Combine human and digital intelligence
Challenge
Achieving fine-tuned
composition of precision-
manufactured parts
requires tight control of
the tools and processes to
ensure consistent quality
and reliability; and
minimize waste.
Strategy
Convert the knowledge
of service technicians
into a digital format and
apply machine learning
to adjust cutting
configuration and
optimize production.
Results
• Retrofit machine learning
information to automatically
adjust equipment and manage
tooling inventory
• Tool-level intelligence to
minimize idle time up to 50%
• Improve Customer service level
“What sets us apart is our deep knowledge of the machining process and our ability to
translate that knowledge into the algorithms used to analyze the data.”
— Mats Lindeblad, Global Product Manager Sandvik Coromant
43. Machine Learning addresses a range of scenarios
High-yield use cases that transform Manufacturing
Supply Chain
and
Inventory
Optimization
Predictive
Maintenance
Anomaly
Detection
Production
Scheduling
Demand
Forecasting
Quality
Assurance
47. Azure IoT Edge
• Low-latency (important in safety scenarios, such as stop the machine immediately)
• Intermittent connectivity (remote mining, oil, factories)
• Integrate sensor data from different devices
• Filter or aggregate telemetry data before sending it to the cloud
• Transform raw input from sensors to meaningful information
On Premise
Azure IoT Edge
Devices
RTOS,Linux,Windows,
Android,iOS
Stream
Analytics
Custom
OPC UA or
Modbus
ingestion
module
Azure
Functions
Machine
Learning
Cognitive
Services
Container
Management
Functions
Runtime
Device Twin
Local
Storage
IoT Hub
Device
Twin
Stream
Analytics
Machine
Learning
Power BI
51. The industry’s most comprehensive portfolio
Azure IoT Suite (PaaS)
Preconfigured solutions for common IoT scenarios
Data and Analytics
Azure Time Series
Insights
Azure
Machine Learning
Cosmos DB
Azure Stream
Analytics
Azure Data Lake
Azure Data Lake
Analytics
Azure HD Insight
Visualization and Integration
Azure Logic Apps
Notification Hubs
Azure Websites
Microsoft Flow
Microsoft
Power BI
Azure Monitor
Azure Active
Directory
Device support
Azure IoT
Device SDK
Azure IoT
certified devices
Security Program for
Azure IoT
Windows 10 IoT
IoT
Edge
Azure IoT Hub
Azure IoT Edge
IoT Hub Device
Provisioning Service
Microsoft Dynamics
Connected Field Service
Microsoft IoT Central
IoT SaaS
Solutions (PaaS)
Technologies (PaaS)
Solutions (SaaS)