Industrial Edge

Portfolio Architecture presentation of Industrial Edge Validated Pattern

Transcript

1. Industrial Edge Stefan Bergstein Sr. Solution Architect June 2022 Portfolio Architecture Ishu Verma Emerging Technology Evangelist William Henry Sr. Distinguished Engineer
2. Boosting manufacturing efficiency and product quality with artificial intelligence and machine learning out to the edge.
3. Tackling the challenges • Software driven production optimization • Real-time insights on factory data • Declarative configuration management for multi-tiered production lines • Leverage big data and ML technology for traceability and analytics
4. Generic view of the architecture Portfolio Architecture
5. Core Data Center Industrial Edge Introduction Secure GitOps and DevOps management across core and factory sites Provides real-time insights from edge to centralized sites Reducing maintenance costs Business Drivers Edge AI Application Container Images Factory Edge Management Edge Management Local Dashboard Sensor Data DevOps Management ML Model Training Sensor Data Stream Data Lake GitOps Management Open Hybrid Cloud Management Secrets
6. Industrial Edge - Logical view
7. Industrial Edge Secrets Management
8. Industrial Edge GitOps
9. APPENDIX WITH ELEMENT DETAILS
10. Detail: Cluster Management Policy based cluster lifecycle management Output to: - Container control plane
11. Detail: Git Repository (Factory) Git Repository for storing: - Source code - Configuration - Manifest Input from: - Data science Output to: - CI/CD pipeline - Data science
12. Detail: ML CI/CD Machine learning continuous integration/continuous development pipeline Input from: - Container tooling - SCM system - Image registry Output to: - Image registry
13. Detail: Image registry (Core Data Center) Container image registry Input from: - CI/CD pipeline - Regitry.redhat.io Output to: - OCP registry
14. Detail: Image registry (Factory) Integrated Kubernetes image registry Input from: - Quay Output to: - Container control plane
15. Detail: Image registry (Cloud) Certified container image from vendors Output to: - CI/CD pipeline
16. Detail: GitOps Controller Policy based cluster and application lifecycle management Output to: - GitOps
17. Detail: Edge GitOps GitOps agent for cluster and application lifecycle management Input from: - Git repo - GitOps controller Output to: - Container control plane
18. Detail: Cluster Management Policy based cluster lifecycle management Output to: - Container control plane
19. Detail: Edge Message Broker Multi-protocol message broker Input from: - Line data server - Container control plane Output to: - MQTT integration - Messaging consumer
20. Detail: Messaging Consumer Messaging consumer application Input from: - Edge message broker Output to: - Anomaly detection - Dashboard
21. Detail: Distributed Streaming (Factory) Pub/sub message broker, data replication Input from: - MQTT integration - Stream processor Output to: - Stream processor - Distributed streaming (cloud)
22. Detail: Distributed Streaming (Core Data Center) Pub/sub message broker Input from: - Distributed streaming (factory) Output to: - Event stream integration
23. Detail: MQTT Integration Kubernetes native integration framework Input from: - Edge message broker Output to: - Distributed streaming
24. Detail: Anomaly Detection Artificial Intelligence/Machine learning model execution with real time data Input from: - Image registry - Messaging consumer Output to: - Dashboard
25. Detail: Dashboard Dashboard to show key performance metrics Input from: - Web server - Messaging consumer
26. Detail: Line Data Server Edge gateway or edge server interacting with IoT devices Input from: - Sensor Output to: - Edge message broker
27. Detail: Shared Storage (Factory) Container native data storage
28. Detail: Data Lake Persistent data store for edge data Input from: - Firewall - Device data Output to: - Device control - Firewall
29. THANK YOU End of public section