EUDR Enforcement Integration Guide

Everything an enforcement agency or platform partner needs to reproduce the EUDR Enforcement Dashboard using the Epoch SCO2 API.

1. Overview & Architecture

The Enforcement Dashboard provides three complementary workflows for EUDR compliance monitoring, exposed as tabs:

1. Plot Screening (default tab) — Upload GeoJSON, CSV, or XLSX plot polygons (or draw them on the map) to screen against Article 9 (deforestation) and Article 10 (procedural / legality) risk. Supports per-plot validation, optional deforestation and Article 10 checks, and submission to batch processing.
2. Producer Facility Screening — Screen facility points (uploaded, map-clicked, or geocoded from an address) for per-facility confidence scoring plus optional deforestation (50 km radius), Article 10 procedural risk, legality / adverse-media, natural-hazard, and infrastructure checks.
3. Country Commodity Assessment — Aggregate view of all assessed producers, grouped by country and commodity. Supports risk-tiered status (High Risk, Medium Risk, Compliant) based on the percentage of non-compliant area. Enables drill-down to individual plot-level results.

All data is fetched from the Epoch SCO2 REST API. The dashboard is a static HTML page — no backend required beyond the API.

2. Authentication

All API requests require a Bearer token. Obtain one by calling the /refresh_token endpoint with user credentials in a JSON body:

POST https://epoch-sco2-api.com/refresh_token
Content-Type: application/json

{ "email": "USER", "password": "PASS" }

Response:

{ "access_token": "eyJhbGciOiJSUzI1NiIs..." }

Include the token in all subsequent requests:

Authorization: Bearer <access_token>

Base URLs:

• Production: https://epoch-sco2-api.com
• Development: https://dev-epoch-sco2-api.com

3. API Endpoints

GET /fetch_supply_shed?stat_type=all&table_extension=eudr&limit=100000

GET /fetch_deforestation_check?aggregate=false&limit=100000&filename=0x49a7...

POST /validate_locations?location_type=plot&include_art9=true&include_art10=true&include_eudr_profile=true&commodity_type=palm_oil&limit=10000
Content-Type: multipart/form-data
Body: file=@plots.geojson

POST /validate_locations?location_type=facility&commodity_type=timber&include_art9=true&include_legality=true&include_eudr_profile=true&check_road_access=true&normalize_input=true
Content-Type: multipart/form-data
Body: file=@facilities.csv

POST /batch_supply_shed
Content-Type: multipart/form-data
Body:
  file=@timber_Suriname_2026-03-11.geojson
  collection_name=timber_Suriname_2026-03-11
  crop_type=timber
  stats=eudr_deforestation
  validate_locations=false
  force_reprocess=false
  eudr_dds=true

GET /export_eudr_dds/?collections=0x49a7...

4. Dashboard Tabs

4.1 DDS Submission Screening (Default Tab)

Endpoint: POST /validate_locations

The default view after login. Upload GeoJSON, CSV, or XLSX files to screen locations. The tab supports two location types:

Input Options

• File upload — Drop a GeoJSON, CSV, or XLSX file. CSV/XLSX files with address columns are automatically geocoded.
• Address geocoding — (Facility mode) Type an address directly. The UI packages it as a CSV, sends it to the API for geocoding via Google, and displays the result on the map.

Configuration

Screening Results

After screening, the sidebar shows summary cards and the map displays screened locations. A data table below the map shows all facilities/plots with sortable columns.

Facility table columns:

Batch Submission

After screening, select facilities/plots via checkboxes and submit to batch processing. The collection name auto-suggests as {commodity}_{country}_{date} and updates on re-screening. Facilities go to /batch_supply_shed, plots go to /batch_process.

4.2 Country Commodity Assessment

Endpoint: GET /fetch_supply_shed?stat_type=all&table_extension=eudr&limit=100000

Loads all producer assessments and displays them in a sortable table with risk-tiered compliance status.

Two sub-views:

• Producer View — One row per assessment (collection_id). Default.
• Group View — Rows grouped by company_name. Each group shows its worst-case status and total NC plots. Expandable/collapsible.

Columns:

Summary cards: Total Producers, Compliant, High Risk, Medium Risk (≤5%), Not Assessable.

4.3 Plot Detail View (Drill-Down)

Triggered when a user clicks a producer row. Fetches plot-level data via:

GET /fetch_deforestation_check?aggregate=false&limit=100000&filename={collection_id}

Shows individual plot polygons on the map with binary compliance status (Compliant / Non-Compliant based on >0.5 ha noncompliance area). Summary cards show Total Plots, Compliant, and Non-Compliant counts.

5. Facility Confidence Scoring

Facility confidence uses a penalty-based model: each facility starts at 1.0 and has points deducted for missing or negative attributes.

Penalty Breakdown

Confidence Levels

Building Density Threshold

A facility is classified as high_density (urban) when building footprint area exceeds 50% of the 100m buffer area around the facility point. This is a negative signal — commodity processing facilities are typically in rural or peri-urban areas, not dense urban centers.

Facility Tooltip

On hover, the map tooltip shows the full scoring breakdown including each attribute's value and its penalty contribution. high_density = Yes should display in red (negative signal), while high_density = No displays in green (positive — rural area).

Producer Name Extraction

The API extracts producer_name from input CSV/XLSX columns by scanning for columns matching keywords: "name", "company", "producer", "facility". Columns that are part of the API's own output (facility_address, facility_id, commodity_type, scoring fields, etc.) are excluded from the scan to prevent false matches (e.g. address leaking into producer name).

6. Risk Classification Logic

Risk status is derived client-side from the API response fields. The key field is noncompliance_area_perc — the fraction of total assessed area flagged as non-compliant.

Producer-Level Risk (Aggregate)

function getComplianceStatus(p) {
  const canDo = isAssessable(p.locations_confidence)
    || p.noncompliance_plot_count != null
    || p.noncompliance_area_ha != null;
  if (!canDo) return { label: 'Not Assessable', cls: 'unknown' };

  const ncPerc = Number(p.noncompliance_area_perc || 0);
  const ncPlots = Number(p.noncompliance_plot_count || 0);
  const ncArea = Number(p.noncompliance_area_ha || 0);

  if (ncPerc > 0.05 || ncPlots > 100) return { label: 'High Risk', cls: 'high-risk' };
  if (ncPerc > 0 || ncPlots > 0 || ncArea > 0)
    return { label: 'Medium Risk', cls: 'medium-risk' };
  return { label: 'Compliant', cls: 'compliant' };
}

Plot-Level Compliance (Detail View)

Plot-level compliance uses a simple binary classification based on the noncompliance_area field (deforested area in hectares). A plot is non-compliant if more than 0.5 ha of deforestation is detected.

// Binary compliance for individual plots
const ncArea = Number(p.noncompliance_area || 0);
const isNonCompliant = ncArea > 0.5;
const status = isNonCompliant ? 'Non-Compliant' : 'Compliant';

Assessment Type

if (area_ha_supply_shed > 0 && stat_type.includes('supply_shed')):
    type = "Supply Shed"    // isochrone-based area assessment
else:
    type = "Direct Plots"   // specific plot polygons provided

Assessability Check

function isAssessable(confidence) {
  return ['high', 'medium', 'low'].includes(confidence?.toLowerCase());
}

// A producer can be assessed if confidence is valid OR noncompliance data exists
function canAssess(p) {
  return isAssessable(p.locations_confidence)
    || p.noncompliance_plot_count != null
    || p.noncompliance_area_ha != null;
}

7. Map Visualization

The POC uses deck.gl with MapLibre GL. Any map library works.

Aggregate View (Country Commodity Assessment)

Facility points colored by risk status. Sort features so higher-risk points render on top (last in the array):

const riskOrder = { 'compliant': 0, 'unknown': 1, 'medium-risk': 2, 'high-risk': 3 };
pointFeatures.sort((a, b) =>
  (riskOrder[getComplianceStatus(a.properties).cls] || 0) -
  (riskOrder[getComplianceStatus(b.properties).cls] || 0)
);

Fill colors:

• High Risk — [220, 53, 69, 200]
• Medium Risk — [202, 138, 4, 200]
• Not Assessable — [148, 163, 184, 200]
• Compliant — [22, 163, 74, 200]

Point coordinates come from the feature's GeoJSON geometry, or parse facility_geo from metadata (WKT point) for facilities without geometry.

Screening View (DDS Submission Screening)

Facility points colored by confidence level:

• High confidence — Green
• Medium confidence — Yellow
• Low confidence — Red

Plot polygons colored by per-plot risk score or filtered by metric (overall risk, deforestation, protected area overlap).

Detail View (Plot Polygons)

Plot polygons from /fetch_deforestation_check, colored by binary compliance (>0.5 ha noncompliance area):

• Non-Compliant — [220, 53, 69] (red)
• Compliant — [22, 163, 74] (green)

Overlay the supply shed boundary (blue outline) from the parent feature's geometry, and mark the facility location (yellow dot) if available.

8. Batch Processing Workflow

The screening-to-processing workflow in the dashboard:

1. Screen — Upload file and run /validate_locations to get confidence scores and validation results.
2. Review — Inspect results in the table and map. Select facilities/plots to submit via checkboxes.
3. Submit — Click "Process Supply Shed" (facilities) or "Process Plots" (plots). This sends selected features to /batch_supply_shed or /batch_process respectively.
4. Monitor — Processing runs asynchronously via Prefect flows. Results appear in the Country Commodity Assessment tab once complete.

Collection Naming

The collection name auto-suggests as {commodity}_{country}_{date} (e.g. timber_Suriname_2026-03-11) based on the first feature's properties. It updates automatically on re-screening.

Batch Parameters

• stats=eudr_deforestation — Always set for EUDR enforcement workflows
• validate_locations=false — Already validated during screening step
• force_reprocess=false — Prevents duplicate processing of existing collections
• eudr_dds=true — (batch_supply_shed only) Generates the DDS audit trail

9. Audit Trail (DDS) Export

Endpoint: GET /export_eudr_dds/?collections={collection_id}

Provide a "Download DDS" button for each assessed producer. Trigger a binary ZIP download. Disable the button for "Not Assessable" rows.

The ZIP contains the full Due Diligence Statement including:

• Per-plot deforestation results with alert level and confidence
• Satellite data references (dates, sensors, resolution)
• Geographic scope and monitoring period
• Compliance determination methodology

10. Parsing NDJSON Responses

Both /fetch_supply_shed and /fetch_deforestation_check return Newline-Delimited JSON (NDJSON). Each line is a complete JSON object (GeoJSON Feature).

JavaScript Example (Streaming)

async function fetchNDJSON(url, token) {
  const resp = await fetch(url, {
    headers: { 'Authorization': `Bearer ${token}` }
  });
  const features = [];
  const reader = resp.body.getReader();
  const decoder = new TextDecoder();
  let buffer = '';
  while (true) {
    const { done, value } = await reader.read();
    if (done) break;
    buffer += decoder.decode(value, { stream: true });
    const lines = buffer.split('\n');
    buffer = lines.pop(); // keep incomplete line in buffer
    for (const line of lines) {
      const trimmed = line.trim();
      if (!trimmed) continue;
      try {
        const obj = JSON.parse(trimmed);
        if (obj.type === 'Feature') features.push(obj);
        else if (obj.type === 'FeatureCollection') features.push(...(obj.features || []));
      } catch (_) {}
    }
  }
  // Process remaining buffer
  if (buffer.trim()) {
    try { const obj = JSON.parse(buffer.trim()); features.push(obj); } catch (_) {}
  }
  return features;
}

Python Example

import requests, json

def fetch_ndjson(url, token):
    resp = requests.get(url, headers={"Authorization": f"Bearer {token}"}, stream=True)
    features = []
    for line in resp.iter_lines(decode_unicode=True):
        if line.strip():
            features.append(json.loads(line))
    return features

Quick Start Checklist

1. Authenticate via /refresh_token — store the Bearer token
2. Screen locations: Upload a file to /validate_locations with location_type=facility or plot
3. Review facility confidence scores and plot validation results
4. Submit selected locations to /batch_supply_shed (facilities) or /batch_process (plots)
5. Monitor results: Fetch all producers via /fetch_supply_shed?stat_type=all&table_extension=eudr&limit=100000
6. Derive risk status client-side using noncompliance_area_perc (see Section 6)
7. Sort map features by risk tier so red points render on top
8. Display summary cards: Compliant, High Risk, Medium Risk (≤5%), Not Assessable
9. On row click: fetch plots via /fetch_deforestation_check?aggregate=false&limit=100000&filename={collection_id}
10. Color plot polygons by binary compliance: red if noncompliance_area > 0.5 ha, green otherwise
11. Download audit trail via /export_eudr_dds/?collections={collection_id}