Build a Sparkplug B Sensor Drift Calibration Rebalancer

Evaluate normalized Sparkplug B dataset metrics against sensor calibration thresholds to detect measurement drift. Synthesize a recalibration and replacement schedule for instrumentation technicians.

Instrumentation Lead

Ingest normalized Sparkplug B device metric records containing calibration dates and drift standard deviations.

Compare sensor variance against published allowable tolerance thresholds.

Generate a prioritized recalibration schedule grouped by line and technician specialization.

Export a simulated maintenance dispatch record for supervisor sign-off.

Terminal dashboard showing sensor drift variance and calendar calibration rosters.

Generate local synthetic JSON mock streams containing normalized Sparkplug B metric records with sensor serials and baseline offsets. Starter examples below are synthetic. They are not downloaded records.

Eclipse Sparkplug Utilizes Sparkplug B metric property metadata, dataset types, and device lifecycle states outlined in Source 2.

pydantic-ai, fastapi. Pydantic AI provides strict schema enforcement on nested metric records and validation of calibration schedules. Local small LLM (e.g., Llama 3 8B) or hosted Claude via API for schedule synthesis.

Submit runnable source, pinned dependencies, setup commands, a working interface, example outputs, and a recorded end-to-end demo. Include tests and an action log.

Correctly flags sensors exceeding tolerance thresholds.

Calculates replacement due dates based on historical calibration intervals.

Emits valid JSON schedule artifacts matching the schema.

```json [ { "input": { "device_id": "EDGE-NODE-01/DEV-TEMP-99", "metric_name": "FurnaceTemp", "last_calibrated_days": 190, "drift_deviation_pct": 4.2, "max_allowed_drift_pct": 2.5 }, "expected_output": { "device_id": "EDGE-NODE-01/DEV-TEMP-99", "action_required": "RECALIBRATE", "priority": "HIGH", "estimated_labor_hours": 1.5, "recommended_window": "NEXT_DOWNTIME_SHIFT" } }, { "input": { "device_id": "EDGE-NODE-02/DEV-PRESS-04", "metric_name": "HydraulicPressure", "last_calibrated_days": 45, "drift_deviation_pct": 0.8, "max_allowed_drift_pct": 3 }, "expected_output": { "device_id": "EDGE-NODE-02/DEV-PRESS-04", "action_required": "MONITOR", "priority": "LOW", "estimated_labor_hours": 0, "recommended_window": "ROUTINE_CYCLE" } } ] ```

Read-only advisory tool; produces draft maintenance calendars without commanding field instrumentation. Use mock external writes. Require human review before any real deployment.