GuardianAI_Overview for Condition Monitoring.pptx

Condition Based Monitoring using Artificial Intelligence

1. FactoryTalk® Data, Edge, and Analytics Platform Comprehensive edge-to-cloud platform driving business outcomes Predictive and prescriptive solutions Soft sensing, anomaly detection, condition monitoring, operational KPIs Configure-and-go ML that works with OT domain expertise AI/ML & Data Science Closed-loop machine learning with targeted use cases Best-in-class consulting and managed services Integrated strategy, software, hardware, implementation & support Leverage learnings from impressive implementation portfolio Closed loop feedback Low-latency control Light footprint compute surface Equipment Connect & control disparate Rockwell Automation & third-party industrial devices OT data contextualization Flexible common information model Relational, time series, big data storage Cloud API data access Data Management Accelerate IT/OT convergence with contextualized industrial data Container / VMs orchestration Centralized device/fleet management Private edge app marketplace Highly secure posture Edge Management Orchestrate and manage edge devices remotely in a flexible manner Interactive & collaborative storyboards Predictive operational KPIs Batch & performance management Workforce training & instructions Applications Frame work inclusive of IIoT, AR, and Visualization

2. DESCRIPTIVE DIAGNOSTIC PREDICTIVE PRESCRIPTIVE DEVICE SYSTEM ENTERPRISE Am I running ok? Why did a fault happen? I predict a fault will happen soon. What action should be taken to avoid the fault? Is Line 1 running ok? Why is Line 1 quality poor? I predict that Line 1 quality is moving out of tolerance. What action should the operator take to avoid poor quality? Which facility performed the best? Why is Site A throughput behind plan? I predict that Site A will be behind plan soon. What action should I take to avoid Site A from falling behind plan? The Analytics Landscape

3. Asset Health Monitoring on the Edge Leverage AI for early anomaly detection Minimize unplanned downtime Monitor critical assets on the plant floor Identify anomalies before they become critical Electrical Signal Drive Data Vibration monitoring Comprehensive condition monitoring Process data Deployed at the edge

4. GuardianAI Combines electrical, vibration, and process data for holistic condition-based monitoring at scale AI using Electrical data Known FP + Domain Expertise ML Anomaly Detection Expert-provided label AI using Vibration Data AI using System-level Data Correlation, Classification, Root-cause Analysis ML Anomaly Detection ML Anomaly Detection for High-Dimension D ata Known FP + Domain Expertise AI using process data Known FP + Domain Expertise ML Anomaly Detection Expert-provided label Continuous Learning Expert Input Continuous Learning Expert Input Expert Input Expert-provided label Continuous Learning

5. Drives (Electrical Data) Vibration Monitoring iTrak / Kinetix . . . Where do we apply GuardianAI now? Optimized for the Rockwell Automation Intelligent Devices Portfolio

6. . Drives (Electrical Data) Where do we apply GuardianAI now? Optimized for the Rockwell Automation Intelligent Devices Portfolio First Product Release Focus

7. Workflow AI maps asset behavior and flags when deviating from normal behavior

8. AI as electrical signal expert Direct integration to RA drives Embedded Expertise Performs Time and Frequency Domain analysis Acquires buffered drive signals via trend object (1024 @ 256µs / 1024 @ 125µs) Data fidelity is enhanced algorithmically Useful in detecting bearing fault, stator fault, broken-bar, misalignment as well as application related faults Anomaly Detection + Continuous Learning AI engine learns a baseline then detects deviation AI engine absorbs expert feedback in the form of labels for identified operation states Integrates with automation & enterprise software: Simplifies scalable deployment Minimizes infrastructure and maintenance costs Optimized for LV and MV Drives 755, 755T, 6000T

9. Understanding the detected anomaly 1) Embedded Expertise (known fault) 2) Classify Anomalies (unknown fault)

10. Test results against drive data Validated against experimental set up and historized data Experimental Data (Bearing Faults) Complex Bearing Fault Ball Faulted Outer Race Fault Inner Race Fault Healthy Bearing Historized Data (Misalignment Detection)

11. Deployment at the Edge Example Assets Drives/Monitoring Control Layer Edge Analytics Dynamix-1444 PowerFlex (755,755T,6000T) Electrical Signature Vibration Signature AI Based Asset Monitoring & Anomaly Detection Anomaly Warning Tags Motors/Pumps Conveyors Fans Gearboxes Accelerometers – Vibration Monitoring Cloud Layer CMMS/ARP FT Edge Deployment / Orchestration Edge Deployment

12. Drives (Electrical Data) Vibration Monitoring iTrak / Kinetix . . . Where do we apply GuardianAI now? Optimized for the Rockwell Automation Intelligent Devices Portfolio

13. Vibration Monitoring . Where do we apply GuardianAI now? Optimized for the Rockwell Automation Intelligent Devices Portfolio

14. AI as vibration expert Integration to Dynamix & PLC Embedded Expertise Performs analysis in both time and frequency domain Acquires buffered data via Dynamix-1444 or PLC ISO standards coupled with machine learning provides immediate value Model for bearings, gearboxes, pumps, conveyors, fans, compressors and more Anomaly Detection + Continuous Learning AI engine learns a baseline then detects deviation AI engine absorbs expert feedback in the form of labels for identified operation states Integrates with automation & enterprise software: Simplifies scalable deployment Minimizes infrastructure and maintenance costs Released under NDA

15. Test results against vibration data Validated against experimental set up Ball Faulted Bearing Inner Race Faulted Bearing Outer Race Faulted Bearing Combination Bearing Good Bearing BSF Health 0.798044 0.744695 0.701616 1.088922 0.581178 BPIR Health 0.877649 1.143909 0.820802 1.535437 0.389954 BPOR Health 0.620176 0.569969 0.848690 0.819386 0.349664 Good Inner Race Faulted Outer Race Faulted Ball Faulted Combination Faulted BPOR BPIR BPOR BSF Released under NDA

16. Drives (Electrical Data) Vibration Monitoring iTrak / Kinetix . . . Where do we apply GuardianAI now? Optimized for the Rockwell Automation Intelligent Devices Portfolio

17. iTrak / Kinetix . Where do we apply GuardianAI now? Optimized for the Rockwell Automation Intelligent Devices Portfolio

18. AI for MF/SF iTRAK Systems Embedded Expertise : Rockwell Automation IP Differentiate between mover and track faults Isolate problematic movers Anomaly Detection Learn an operational baseline then monitor deviation from baseline Capture new fault / operating condition signatures Continuous Learning End-user contextualizes new fault / operating condition signature Contextualized signature is automatically embedded Integrates with Automation and Enterprise Software

19. Asset Health Monitoring on the Edge Leverage AI for early anomaly detection Minimize unplanned downtime Monitor critical assets on the plant floor Identify anomalies before they become critical Electrical Signal Drive Data Vibration monitoring Comprehensive condition monitoring Process data Deployed at the edge