2. Agenda
• The Internet of Things
• IoT Solutions
• IoT Solutions with Machine Learning
• IoT Edge Computing
• AI Platform – Deep Learning and Cognitive Services
• Computer Vision and Mix Reality(AR)
• AIoT Use Cases Sharing
2
7. Solution PortalProvisioning API
Identity & Registry Stores
Stream Event Processor
Analytics/
Machine
Learning
Data
Visualization &
Presentation
Device State Store
Gateway/
Edge
Storage
IP capable
devices
Existing IoT
devices
Low power
devices
PresentationDevice and Event Processing
Data
Transport
Devices and
Data Sources
Cloud
Gate-
way
Agent
Libs
Agent
Libs
Control System Worker Role
Agent
Libs
8. Azure Time Series
Insights
Azure Machine
Learning
Azure Stream
Analytics
Cosmos DB Azure Data Lake
Azure Data Lake
Analytics
Azure HD Insight
Spark, Storm,
Kafka
Azure Event Hubs
Microsoft Flow
Azure Logic Apps
Notification Hubs
Azure Websites
Microsoft Power
BI
Azure Active
Directory
Azure IoT Hub
Azure IoT Hub
Device Provisioning
Service
Azure IoT Edge
Azure Monitor
PaaSServices&
DeviceSupport
Edge Support
Device Support
Azure IoT Device
SDK
Certified Devices
Azure Certified for
IoT
Security Program
for Azure IoT
IoT Services Data & Analytics Services Visualization & Integration Services
IoTSolutions
(PaaS)
IoTSolutions
(SaaS)
Microsoft IoT Central
IoT SaaS
Microsoft Connected Field Service
Field Service SaaS
Remote Monitoring Predictive Maintenance Connected factory
Windows 10 IoT
Core
Azure IoT Suite
9. 使用 Azure IoT 裝置 SDK 來實作用戶端應用程式,
以便在裝置硬體平臺與作業系統上執行。
裝置 SDK 包含程式庫,可協助將遙測傳送至 IoT
中樞,並接收雲端到裝置訊息。 當您使用裝置
SDK 時,您可以從數種網路通訊協定中選擇,以便
與 IoT 中樞通訊
ConnectivityThings
•Azure IoT SDK for C
•Azure IoT SDK for Python
•Azure IoT SDK for Node.js
•Azure IoT SDK for Java
•Azure IoT SDK for .NET
Windows 10 IoT Core
Windows 10 IoT Enterprise
Linux
Android
RTOS’s (Real Time Operating System)
(ex. VxWorks)
Many others-BLE ,Modbus RTU ,LoRa…
IP connectivity
Built in security
10.
11. 從設備端到資料洞察到採取行動,改變整個企業,走向全世界
架設在行業領先的雲平臺之上
安全
端到端
覆蓋設備端、網路連接、資料加
密以及雲端
開放
連接任意物件
相容任意設備、作業系統、資料
來源、軟體和服務
快速
數分鐘內可開展
預置的方案範本適用於大多數的
物聯網場景
擁有全球領先的商業智慧和分析平臺*
彈性
對高增⾧應付自如
支援數百萬的設備、TB級別的資
料、本地或雲端部署,輕易覆蓋
全球30多個地區
人數據 洞察 行動閘道設備
*February 2015. The Gartner Magic Quadrant for Business Intelligence and Analytics Platforms is the property of Gartner, Inc. and available upon request from Microsoft. Gartner does not endorse any vendor, product or service depicted in its research publications, and
does not advise technology users to select only those vendors with the highest ratings or other designation. Gartner research publications consist of the opinions of Gartner's research organization and should not be construed as statements of fact. Gartner disclaims all
warranties, expressed or implied, with respect to this research, including any warranties of merchantability or fitness for a particular purpose. The above graphics were published by Gartner, Inc. as part of a larger research document and should be evaluated in the
context of the entire document.
23. Bringing the best of AI to Azure and the best of Azure to AI
Microsoft AI Platform
AI Services
AI Infrastructure
AI Tools
PRE-BUILT AI CONVERSATIONAL AI CUSTOM AI
Cognitive Services Bot Framework Azure Machine Learning
AI ON DATA AI COMPUTE
Data
Lake
SQL
Server
Cosmos
DB
Spark DSVM Batch AI AkS
Azure ML
Studio
Azure
Notebooks
VS Tools for
AI/AML
DEEP LEARNING FRAMEWORKS
Cognitive
Toolkit
TensorFlow Caffe2
Others (Azure Workbench, Pycharm…)
Others (Scikit-learn, Keras, PyTorch, MxNet, Chainer…)
CODING AND MANAGEMENT TOOLS
IoT
AI SILICON
24.
25. A variety of real-world applications
Vision Speech
Intent: PlayCall
Language Knowledge Search
26.
27. Bringing the best of AI to Azure and the best of Azure to AI
Microsoft AI Platform
AI Services
AI Infrastructure
AI Tools
PRE-BUILT AI CONVERSATIONAL AI CUSTOM AI
Cognitive Services Bot Framework Azure Machine Learning
AI ON DATA AI COMPUTE
Data
Lake
SQL
Server
Cosmos
DB
Spark DSVM Batch AI AkS
Azure ML
Studio
Azure
Notebooks
VS Tools for
AI/AML
DEEP LEARNING FRAMEWORKS
Cognitive
Toolkit
TensorFlow Caffe2
Others (Azure Workbench, Pycharm…)
Others (Scikit-learn, Keras, PyTorch, MxNet, Chainer…)
CODING AND MANAGEMENT TOOLS
IoT
AI SILICON
28. Building your own AI models for Transforming Data into Intelligence
Prepare Data Build & Train Deploy
29. Lifecycle 1. define problem
2. acquire + process data
3. design model architecture
4. train model5. test/evaluate
a. initialize
b. feed in minibatch of data
c. calculate lossd. optimize: minimize loss
e. update weights
y =Wx + b
loss = |desired – actual outcome|δ
6. deploy
30. Prepare
Data
Register and
Manage Model
Train & Test
Model
Build
Image
Build model
Azure Notebooks
Deploy
Service
Monitor
Model
Prepare Experiment Deploy
Azure Machine Learning Process
31. Machine Learning Studio 互動式工作區
提供互動式的視覺化工作區,讓您輕鬆建置、測試和反覆運算預測分析模型。 您可以將「資料集」和分析「模組」拖放到
互動式畫布,將它們連接在㇐起以構成「實驗」,然後在 Machine Learning Studio 中執行
32. Anomaly detection models in Azure Stream Analytics
Built-in ML models for anomaly detection in Azure Stream Analytics significantly
reduces the complexity and costs associated with building and training machine
learning models.
•AnomalyDetection_SpikeAndDip function to detect
temporary or short-lasting anomalies such as spike or
dips.
•AnomalyDetection_ChangePoint function to detect
persistent or long-lasting anomalies such as bi-level
changes, slow increasing and slow decreasing trends.
SELECT sensorid, System.Timestamp as time, temperature as temp,
AnomalyDetection_SpikeAndDip(temperature, 95, 120, 'spikesanddips’)
OVER PARTITION BY sensorid
LIMIT DURATION(second, 120) as SpikeAndDipScores
FROM input
41. The vision AI developer kit
- Qualcomm® Technologies, Inc. and Microsoft
collaboration
- Run AI models on the edge without additional
computers or web connection or leverage the cloud
- Create, deploy and manage all your models in the cloud
and the edge with Azure ML and Azure IoT Edge
- Register for early access preview today:
http://www.visionaidevkit.com
45. Azure Spatial Anchors overview
Azure Spatial Anchors empowers developers with essential capabilities to build spatially aware mixed reality
applications. It enables developers to work with mixed reality platforms to perceive spaces, designate precise
points of interest, and to recall those points of interest from supported devices. These precise points of
interest are referred to as Spatial Anchors. It is composed of a managed service and client SDKs for
supported device platforms
• Support Microsoft HoloLens, iOS-based devices
supporting ARKit, and Android-based devices
supporting ARCore.
• Support Unity for creating and deploying mixed reality
applications
• Multi-user experiences. Spatial Anchors makes it easy for
people in the same place to participate in multi-user
mixed reality applications.
• Way-finding. Developers can also connect Spatial
Anchors together creating relationships between them.
• Persisting virtual content in the real-world. An app can
let a user place a virtual calendar on a conference room
wall, that people can see using a phone app or a
HoloLens device.
46. 46
With Azure Spatial Anchors , You can
• Create and locate anchors using Azure Spatial Anchors
• Anchor relationships and way-finding in Azure Spatial Anchors
• Persisting virtual content in the real-world.
• Logging and diagnostics in Azure Spatial Anchors
• Android SDK reference
47. Solution Architecture
Devices Azure
Spatial Anchors with Speech Service Demo
1. 搜尋附近可用錨點
2. 上傳語音並轉換為文字(Cognitive Service)
3. 判別所選物件並產生錨點(上傳至Spatial Anchors)
4. 將產生的錨點和物件儲存至其 Web服務(API)
使用者掃描現場環境, 找到真實世界的可用錨點, 並且
使用Android App透過語音輸入所要物件
Spatial Anchors
Azure App Service
Speech to Text(Cognitive Service)
1
2
Android
Mobile Device
3
4
48.
49. AIoT Use Cases Sharing&Demo
• Sketch2Code
• GAN Network
• Industrial AOI for Laser-cut machinery
• Vision-Guided AGV/Drone
• OpenPose/Posenet on IoT Edge
50.
51. Transform Your handwritten design from a picture to valid HTML markup
https://sketch2code.azurewebsites.net/
52. •A Microsoft Custom Vision Model:
This model has been trained with images of
different HTML elements like buttons, text
box, and images.
•A Microsoft Computer Vision Service:
To identify the text written into a design
element a Computer Vision Service is used.
•An Azure Blob Storage:
All steps involved in the HTML generation
process are stored, including the original
image, prediction results and layout
grouping information.
•An Azure Function:
Serves as the backend entry point that
coordinates the generation process by
interacting with all the services.
•An Azure website:
User font-end to enable uploading a new
design and see the generated HTML results.
54. https://drawingbot.azurewebsites.net/
At the core of Microsoft’s drawing bot is a
technology known as a Generative Adversarial
Network, or GAN. The network consists of two
machine learning models, one that generates
images from text descriptions and another,
known as a discriminator, that uses text
descriptions to judge the authenticity of
generated images
62. • Vision-guided AGVs to
follow a trained object
without human
assistance
• Training on the fly ,
Execution on the edge.
• Dynamic change
object model without
prebuilt cost
• Multi-object detection
for path routing
decision
• Companion Edge
63. • Quick Dashboard for
Drone Flight Statistics
with IoT Central
• Training on the fly ,
Execution on the edge.
• Multi-object detection
for drone control
decision
• Intelligent Edge