Advanced IoT Development · Advanced
Edge Analytics: Rules Engines on Device
Run lightweight scoring at the edge with bounded CPU budgets.
Overview
Move filtering closer to sensors using deterministic rule packs, shadow deployments, and offline caches. You compare batch vs streaming mini-pipelines with measurable watt budgets.
Included focus areas
- CPU budget worksheets per pipeline stage
- Shadow deployment pattern with kill switch
- Offline cache sizing exercises
- Deterministic seeding for reproducible demos
- Heatmaps for sensor fusion conflicts
- Observability hooks without shipping PII
- Edge container lifecycle drills
Outcomes you can evidence
- Publish rule packs with signed manifests
- Demonstrate graceful degradation when uplink drops
- Present watt comparisons across two firmware builds
FAQ
GPU acceleration?
Not covered—labs stay MCU-class with documented CPU ceilings.
Cloud credits?
No cloud credits are bundled; sandbox budgets are classroom-sized only.
Certification?
Completion letters available; third-party exam vouchers are not included.
Participant notes
“Shadow deploy lab mirrored how we stage firmware today—minus the GPU rabbit holes.”
“Kill switch drill felt rushed on Tuesday, but office hours clarified the rollback path.”