Summary
Add an end-to-end vision-based leak detection pipeline blueprint to the edge-ai accelerator. This integrates camera ingestion, AI inference, video capture, edge-to-cloud dataflow routing, and Microsoft Teams notification into a deployable scenario.
Motivation
The edge-ai accelerator needs a concrete, end-to-end scenario that demonstrates the full pipeline from edge camera ingestion through AI inference to cloud-side alerting. Leak detection in industrial settings (oil & gas, utilities, manufacturing) is a high-value use case that exercises all major accelerator components.
Scope
New Features
- Leak detection blueprint — orchestrates components 508 (media-connector), 507 (ai-inference), 503 (media-capture), 130 (messaging dataflows), 045 (notification), and 040 (messaging) into a complete pipeline
- 045-notification Teams integration — Logic App deduplication of alerts and posting to Microsoft Teams channels
- 040-messaging Azure Functions output — expose function app endpoint for notification integration
- CI/CD scripts —
build-leak-detection-images.sh and deploy-leak-detection-apps.sh for automated build and deployment
leak-detection.tfvars.example — example variable file for the full-single-node-cluster blueprint
Documentation
- Scenario guide — step-by-step deployment walkthrough (
docs/getting-started/leak-detection-scenario.md)
- Architecture Decision Record — end-to-end pipeline architecture ADR (
docs/solution-adr-library/leak-detection-e2e-pipeline-architecture.md)
Build & Maintenance
- Shell script reformatting to 4-space indent (shfmt compliance)
- Cargo.lock CVE patches (openssl, rand, rustls-webpki)
- Asset schema pattern fixes and API version alignment to 2026-04-01
- Ruff lint, cspell, and megalinter fixes
- Release workflow updates (GitVersion, release-branch-create pipeline)
Components Affected
src/000-cloud/040-messaging — new Azure Functions outputssrc/000-cloud/045-notification — Teams channel posting supportsrc/100-edge/111-assets — asset schema pattern fixessrc/501-ci-cd/scripts — new leak detection CI/CD scriptsblueprints/full-single-node-cluster — leak detection integration- Multiple shell scripts — shfmt 4-space indent standardization
Summary
Add an end-to-end vision-based leak detection pipeline blueprint to the edge-ai accelerator. This integrates camera ingestion, AI inference, video capture, edge-to-cloud dataflow routing, and Microsoft Teams notification into a deployable scenario.
Motivation
The edge-ai accelerator needs a concrete, end-to-end scenario that demonstrates the full pipeline from edge camera ingestion through AI inference to cloud-side alerting. Leak detection in industrial settings (oil & gas, utilities, manufacturing) is a high-value use case that exercises all major accelerator components.
Scope
New Features
build-leak-detection-images.shanddeploy-leak-detection-apps.shfor automated build and deploymentleak-detection.tfvars.example— example variable file for the full-single-node-cluster blueprintDocumentation
docs/getting-started/leak-detection-scenario.md)docs/solution-adr-library/leak-detection-e2e-pipeline-architecture.md)Build & Maintenance
Components Affected
src/000-cloud/040-messaging— new Azure Functions outputssrc/000-cloud/045-notification— Teams channel posting supportsrc/100-edge/111-assets— asset schema pattern fixessrc/501-ci-cd/scripts— new leak detection CI/CD scriptsblueprints/full-single-node-cluster— leak detection integration