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

InputDescription
Yield point layerRaw point data from monitor export.
Yield fieldField containing raw yield values.
Optional telemetry fieldsSpeed and heading fields for telemetry QA.
Optional moisture fieldMoisture field for dry-yield normalization.

Key Settings

SettingDefaultGuidance
output_prefixrequiredPrefix used for all emitted artifacts.
profilebalancedUse fast, balanced, or strict.
keep_intermediatestrueKeep intermediate branch outputs for review.
filtering_modeprofile-basedstandard or robust.
lag_correction_modenoneSet distance only when lag distance is known.
target_moisture_pct15.5Used only if a moisture field is supplied.

What You Get

DeliverableFormatDescription
qa_flagsGeoPackageEdge QA points.
clean_pointsGeoPackageFinal normalized points.
clean_mapGeoPackageFinal swath map polygons.
confidence_pointsGeoPackageFinal points with confidence field (QA_CONF).
summaryJSONRun summary and branch diagnostics.
html_reportHTMLOptional 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_confidence and 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.
  • precision_ag_yield_zone_intelligence
  • field_trafficability_and_operation_planning
  • precision_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.