Conclusion and Next Steps

What You’ve Accomplished

Congratulations! You’ve completed the AI Lifecycle at the Edge workshop. You’ve explored a complete solution spanning edge devices and cloud-like infrastructure:

Module 1 - Transportation robot:

  • Connected to robot infrastructure via MicroShift

  • Deployed MinIO object storage for AI models

  • Configured model serving with KServe/OpenVINO

  • Deployed Battery Monitoring System with real-time predictions

Module 2 - Red Hat OpenShift AI Configuration:

  • Reviewed Red Hat OpenShift AI installation optimized for edge

  • Explored data science project organization

  • Examined data connections linking SNO to robot storage

Module 3 - Model training:

  • Reviewed JupyterLab workbench configuration

  • Imported training notebooks from GitHub

  • Trained Stress Detection and Time-to-Failure models

Module 4 - Model serving:

  • Reviewed OpenVINO model servers on SNO

  • Understood KServe v2 inference API protocol

  • Queried endpoints to validate predictions

Module 5 - Pipeline automation:

  • Reviewed pipeline server architecture

  • Executed automated retraining pipeline

  • Scheduled pipelines for continuous model improvement

Module 6 - Test alerts:

  • Tested battery health alerts dashboard

  • Verified auto-inference and AI predictions

  • Simulated battery stress scenarios

  • Confirmed alerting system functionality

What You’ve Learned

✓ Edge AI is practical - MicroShift and SNO enable AI at resource-constrained locations

✓ Automation is essential - Pipelines eliminate manual intervention for edge deployments

✓ MLOps at the edge works - Complete ML lifecycle (train → serve → monitor → retrain)

✓ Red Hat provides the platform - OpenShift AI integrates the full AI/ML toolchain

✓ Real-world value - Factory robots operate more efficiently with AI-powered battery monitoring

Next steps

References

Thank you for participating in this workshop!