Platform
The intended field-to-decision system.
Swathmark is being built to connect operational field imagery, biological threshold policies, and decision traceability. This page describes the intended product architecture. For where the product stands today, get in touch — we walk qualified farms, researchers, and equipment teams through it directly.
Stage 1
Capture
Operational field imagery starts with registration: the farm, field, crop, capture session, camera rig, calibration, and operating geometry are recorded so every asset arrives with the metadata a defensible decision requires.
Capture session
Farm, field, rig, capture geometry, rights status, and quality status.
Stage 2
Verify
Before an asset enters the learning workflow, the platform is designed to validate data rights, provenance, metadata completeness, calibration status, and image quality. Training eligibility is intended to be an enforceable system state.
Stage 3
Detect
Versioned computer-vision models are applied to identify relevant pests, life stages, and spatial observations. Uncertain results are flagged for review rather than silently accepted.
- Species
- Colorado potato beetle
- Life stage
- Larva
- Detector
- Versioned model
- Review
- Flagged for expert
Detection review
Field image, detection overlays, life stage, uncertainty, and review state.
Stage 4
Decide
Model outputs, biological context, and uncertainty are evaluated against a versioned threshold policy. The intended outcome is an explicit treatment, no-treatment, or abstain decision — with the reasoning retained.
Biological summary
Qualitative density, life stage, crop stage, and evidence completeness.
Stage 5
Export
The decision is represented as a geographic treatment zone and machine-readable decision record, designed for a future third-party equipment interface.
Treatment zone
Generic field geometry with treatment and no-treatment areas and export type.
Decision trace
Traceability across the whole chain.
Swathmark is being designed so a treatment decision can be traced to its source media, field context, rights status, annotations, dataset, model, calibration, threshold policy, review history, and exported treatment zone.
- Agreement
- Farm
- Field
- Capture session
- Raw media
- Annotation
- Dataset
- Model
- Calibration
- Threshold policy
- Treatment decision
- Export
Decision trace
Lineage chain from agreement and capture through model, policy, decision, and export.
Product boundaries
What the platform is not.
- Swathmark is not a dedicated treatment machine, and dedicated treatment machinery is not part of the current platform.
- Swathmark is not a generic farm-management dashboard.
- Swathmark does not provide agronomic, pesticide-label, or treatment advice.
- Platform extension beyond Colorado potato beetle in potatoes is a design direction, not a claim of current production performance.
In depth
Each stage, in detail.
Each page states its inputs, outputs, limitations, and what has and has not been demonstrated.
- Agricultural computer visionWhat agricultural computer vision estimates from field imagery, what it cannot establish on its own, and why a detection is not yet a treatment decision.
- Pest decision platformHow a pest decision platform differs from pest detection: the verified inputs it requires, the treat, no-treat, or abstain outputs it produces, and who uses it.
- Decision traceabilityWhat decision traceability means in practice: the provenance, model and policy versions, review history, and decision records needed to reconstruct a treatment decision.
- Biological threshold policiesWhat a biological threshold policy contains, how it differs from a detection, who sets thresholds, and why policy is versioned separately from model code.
- Uncertainty and abstentionWhy an agricultural decision system must be able to abstain, when it should, how uncertainty differs from abstention, and what happens after an abstain result.
- Field data verificationThe capture metadata, rights, calibration, and image-quality checks a field image must pass before it can support a decision or enter a training dataset.
- OEM integrationHow an equipment manufacturer would consume a Swathmark decision: interface design, the edge and cloud boundary, treatment-zone exports, and compatibility status.
- Colorado potato beetleColorado potato beetle biology and life stages, what computer vision can detect for each, why Swathmark starts here, and the current validation and evidence status.
- GlossaryThe definitions used consistently across every page of this site.