IoTSummit: Create iot devices connected or on the edge using ai and ml
1 like105 views
This document summarizes an IoT presentation about Azure IoT Edge. It discusses Azure IoT Edge's capabilities including running AI models and containers at the edge, deploying cognitive services containers, adding resiliency with Kubernetes, and monitoring edge devices. It also previews new IoT Edge certified edge servers and gateways from Nvidia and demonstrates logging device data in real-time.
IoTSummit: Create iot devices connected or on the edge using ai and ml
- 1. September 14 Create IoT devices connected or on the Edge Using AI and ML
- 2. About Me • Marco Dal Pino o Enterprise IT Manager o Microsoft MVP o Intel BlackBelt & Innovator • Business Interests o IoT, IIoT, Embedded System o MR/AR/VR Virtual vision in general o AI / Cognitive services o Cloud (Azure) • Contact o About.me/marcodalpino
- 3. Thanks to our Sponsors: With the support of:
- 5. Value enabled by IoT from B2B scenarios McKinsey and Company 70% 20B 20 billion connected devices by 2020 Gartner IoT is enabling the digital transformation of every business on the planet $12.4B 180 180 zettabytes of digital data by 2025 IDC Industrial IoT Platforms for Manufacturing market to become a $12.4B opportunity by 2024, driven by deployments in discrete manufacturing environments IoT Analytics
- 6. Microsoft is investing 5 billion dollars in IoT over the next 4 years
- 8. Microcontroller Azure Sphere • Integrated Circuit designed to govern a specific operation in an embedded system • Highly-secured, connected MCU • Azure Sphere Linux OS for modern MCUs • Included Azure IoT Device SDK IoT Devices Azure IoT Device SDK • Endpoint devices such as appliances, vehicles, or factory machines that connect, interact and exchange data • 1000+ devices, 250+ partners - all certified to work great with Azure IoT Services • Cross-platform and open source: Windows IoT, Linux, Android, iOS, RTOSs and more Edge Devices Azure IoT Edge • Devices that aggregate, process & provide gateway capabilities for IoT endpoints • Deploy and manage Azure Services in containers on any IoT device • AI, AzureML, Azure Stream Analytics and more • Cross-platform and open source: Windows IoT, Linux Edge Appliances Azure Data Box Edge • Integrated appliances that provide a subset of cloud edge roles, such as ML- inferencing • Data Box Edge: AI-Enabled, Storage and compute Azure Edge appliance • Data Box: Offline, ruggedized data transport, 100 TB – 1 PB Edge Stack Azure Stack • Scalable solutions that provide a full cloud stack, including IaaS and PaaS capabilities • Edge and Disconnected Scenarios • Regulatory Requirements • Cloud app model on- premises Hyperscale Cloud Edge Regions • First-party cloud regions • Full Range Hyperscale Cloud Services • Tiered Service availability: Heroes > Hubs > Satellites • Open Source Based Services & Tools Most specialization Fewest services Fewest form factors Most services © Microsoft Corporation Sensors + Control Sensors to Interactive Integrated Platform Global scale processing
- 10. Ecosystem momentum System Integrators & Advisors Solution Providers Solution Aggregators Devices IHV Device Connectivity Provider ISV Service Provider SI Professional Services Consulting Solutions Customer
- 13. Things IoT application pattern + edge intelligence Insights Actions Cloud Gateway Insights Actions Azure IoT Hub
- 14. Edge intelligence enabled with Azure IoT Edge Azure IoT Hub Custom Code (data transform) Container Registry Container Video Camera Custom Code (action) Container Azure Cognitive Services (insight) Container IoT Edge runtime Workload description
- 15. Basic Azure IoT Edge functionality
- 16. Azure IoT Edge security Trusted Execution Env Secure Processor Hardware security module Module Module Module Secure Boot and Updates ROM Sensitiv e code IoT Edge runtime Security monitor
- 18. Sensors Constrained Accelerated Enabling intelligent edge spectrum Industrial Server Stack NOW GA on Windows IoT Enterprise!!
- 20. IoT Edge deployments are growing 40% each month since general availability 50+ IoT Edge devices have been certified and available in our Azure Certified for IoT Device Catalog Great engagement with the community on our IoT Edge open source GitHub repository As a leading global energy company, it is imperative we are innovating with the latest technology to improve worker safety and reduce environmental impact – central to our core values. Edge computing is the next wave of cloud innovation. With Azure IoT Edge, we are able to create intelligent applications that can run right on the device and in remote areas with limited or interrupted connectivity, increasing uptime and real-time insights on performance of equipment.
- 23. AI on IoT edge updates
- 24. Telemetry • … Camera as a SensorSpeech Predictive maintenance Predictive maintenanceAsset configuration Smart retail Workplace safety Smart city
- 25. • Small footprint, optimized for Edge with built-in AI Choice of platform AI built-inTime-series built-in Small-footprint engine <500MB Built-In Time-Series Streaming and Analytics Machine learning for low- latency analytics Turn-key management Single management plane to deploy & update Unparalleled performance and security Most secure with industry leading performance SQL Time- Series 0 20 40 60 80 100 120 140 160 180 200 Vulnerabilities(2010-2018) Arm64 x64 Join the Early Adopter Program: http://aka.ms/sqldbe
- 26. • Online Cortana skills • Your own bot, offline http://aka.ms/CortanaSDK http://aka.ms/speechcontainerspreview
- 27. Cognitive Services deployment from the Azure Marketplace Azure IoT HubIoT Edge Device (VM) Marketplace Container Registry Deployment Manifest Cognitive Service Pre-built Container View results
- 28. More containers from Cognitive Services http://aka.ms/visioncontainerspreview Now supports exporting object detection models!
- 30. AI Capture & Share Capture AI + • Simplicity • Modularity • Reusability (Edge/Cloud) - • Performances ex • Cognitive Services + • Performances • Modularity - • Complexity ex • Nvidia Deepstream HTTP MQTT Share MQTT
- 31. DEMO Nvidia Deepstream on IoT Edge demo
- 36. Coming this summer… ThinkSystem SE350 Edge Server, EPC300 Edge Gateway & IoT510-C Camera will be IoT Edge certified in the Azure IoT Device catalog Nvidia Deepstream will be available in the IoT Edge marketplace Nvidia Deepstream By Nvidia Nvidia Deepstream is an application framework features hardware- accelerated building blocks that bring deep neural networks and other complex processing tasks into a stream processing pipeline
- 37. • Telemetry (AI model output): o Store and forward upstream via the edgeHub o Bring your own database to store and query locally • Large files (images, videos): o Use a local blob store (same REST API as the cloud) • Now with D2C tiering and time-to-live capabilities • Will be GA soon on Linux & Windows o Use a storage gateway like Databox Edge • Automatically syncs your data
- 39. • Adds hardware failure resilience to IoT Edge deployments by leveraging Kubernetes platform features • Manage applications from IoT Hub with the same familiar app model • Automatic translation to Kubernetes native application model (pods, services, deployments…) • Cluster can be shared by the multiple edge devices aka.ms/iotedge-on-kubernetes
- 40. • Real-time logs from IoT Edge device • End-to-end security brokered and guaranteed by IoT Hub • Zero config on IoT Edge device - just works! aka.ms/iothub-device-streams
- 41. DEMO Device logs in real-time!
- 44. Azure IoT Edge resources • Documentation – docs.microsoft.com • https://docs.microsoft.com/en-us/azure/iot-edge • Bugs – IoT Edge OSS project GitHub issue page • https://github.com/azure/iotedge/issues • Feature requests - Azure IoT Edge User Voice forum • https://feedback.azure.com/forums/907045-azure-iot-edge
- 45. Thanks to our Sponsors: With the support of:
- 46. Upcoming Events SQLSaturday #878 (Sofia), October 12 http://www.sqlsaturday.com/878/ JS Talks (Sofia), November 23 http://jstalks.net/ Microsoft 365 Developer Bootcamp, December 14 http://aka.ms/M365DevBootcamp
- 47. About Me • Marco Dal Pino o Enterprise IT Manager o Microsoft MVP o Intel BlackBelt & Innovator • Business Interests o IoT, IIoT, Embedded System o MR/AR/VR Virtual vision in general o AI / Cognitive services o Cloud (Azure) • Contact o About.me/marcodalpino