Yield Data Conditioning and QA
What This Tool Does
Yield Data Conditioning and QA runs a full yield-cleaning pipeline and produces cleaned points, a cleaned map, confidence points, and a summary report.
Typical Questions This Tool Helps Answer
- Are the raw yield monitor data from this harvest clean enough to use for management zone analysis, field benchmarking, and agronomic reporting?
- Which passes or spatial areas contain yield artifacts from machine start-stop, lag effects, or speed anomalies that would distort zone-level statistics?
- After conditioning and outlier removal, what is the spatial pattern of yield variability and which QA flags remain for agronomist review?
When To Use
- End-of-season yield monitor cleanup
- Multi-machine or multi-header harmonization
- Production QA before agronomic analytics
What You Need
| Input | Description |
|---|---|
| Yield point layer | Raw point data from monitor export. |
| Yield field | Field containing raw yield values. |
| Optional telemetry fields | Speed and heading fields for telemetry QA. |
| Optional moisture field | Moisture field for dry-yield normalization. |
Key Settings
| Setting | Default | Guidance |
|---|---|---|
output_prefix | required | Prefix used for all emitted artifacts. |
profile | balanced | Use fast, balanced, or strict. |
keep_intermediates | true | Keep intermediate branch outputs for review. |
filtering_mode | profile-based | standard or robust. |
lag_correction_mode | none | Set distance only when lag distance is known. |
target_moisture_pct | 15.5 | Used only if a moisture field is supplied. |
What You Get
| Deliverable | Format | Description |
|---|---|---|
qa_flags | GeoPackage | Edge QA points. |
clean_points | GeoPackage | Final normalized points. |
clean_map | GeoPackage | Final swath map polygons. |
confidence_points | GeoPackage | Final points with confidence field (QA_CONF). |
summary | JSON | Run summary and branch diagnostics. |
html_report | HTML | Optional report page. |
Depending on settings, intermediate keys may also be emitted (for example pass_lines, pass_points, filtered_points, reconciled_points).
Runtime Output Keys
result.outputs["qa_flags"]
result.outputs["clean_points"]
result.outputs["clean_map"]
result.outputs["confidence_points"]
result.outputs["summary"]
result.outputs["html_report"]
Common Questions
Q: Which QA metrics should I review first?
A: Start with summary.mean_confidence, summary.telemetry_points_removed, and summary.clean_points_no_edges to assess quality improvement versus retention.
Q: What is a common interpretation mistake? A: Assuming reduced point count means failure; it often indicates successful outlier and edge-noise removal.
Q: Which settings most change final outputs?
A: Branch controls for telemetry QA, lag correction, moisture normalization, robust filtering, and keep_intermediates usually drive the largest differences.
Q: How should operations use the outputs?
A: Use confidence_points (QA_CONF) and clean_map for downstream mapping, and keep qa_flags plus intermediates for audit traceability.
Results Delivery Checklist
- Yield field mapping and alias resolution were confirmed
- Pass reconstruction output was reviewed
- Final cleaned points and map were reviewed
- Summary counts and mean confidence were checked
Operational Notes
- Keep branch settings (
telemetry QA,lag correction,moisture normalization,robust filtering) explicit in delivery notes because they materially change retained-point counts. - Review
mean_confidenceand retained-point metrics together before approving downstream zone analytics. - Retain intermediates for governance-heavy programs; they are the clearest evidence of why records were removed or adjusted.
Related Tools
precision_ag_yield_zone_intelligencefield_trafficability_and_operation_planningprecision_irrigation_optimization
References
- Runtime implementation:
wbtools_pro/src/tools/workflow_products/yield_data_conditioning_and_qa.rs - Precision Agriculture Intelligence bundle overview:
manual/pro-tools-customer/src/precision_agriculture/overview.md
When To Use This Workflow
Use Yield Data Conditioning and QA when you need an auditable cleaning pipeline before yield zoning, benchmarking, and prescription-support analytics.