Smallholder and medium-scale producers of coffee (Coffea arabica / canephora), cacao (Theobroma cacao), and banana (Musa spp.) in Central America frequently work in mountainous or rural plots with intermittent or zero cellular connectivity. Identifying pathogens early—such as coffee leaf rust (Hemileia vastatrix), cacao frosty pod rot (Moniliophthora roreri), or banana black Sigatoka (Pseudocercospora fijiensis)—is critical to timing cultural controls and preventing yield loss, yet cloud-only diagnostic tools fail in the field when connectivity drops.

HealthyCropping (com.keylorpineda.healthycropping) is an open-source, non-commercial, offline-first Android application. It runs float16 TensorFlow Lite models directly on the phone and stores scans, farms and plots locally in SQLite, so diagnosis works without a signal. Optional sync through a FastAPI backend and a Gemini RAG advisory layer grounded in official agronomy manuals (ICAFE, CATIE, CORBANA) are being built.

Agronomic Disclaimer: HealthyCropping is a non-commercial, open-source decision-support tool designed initially for Costa Rica. It supports field decisions and does not replace a certified agronomist or phytosanitary laboratory. Always consult an agricultural professional before applying agrochemicals or removing plants.

Work follows the phases in docs/PLAN.md:

flowchart LR
    A["1. Select Crop\n(Coffee / Cacao / Banana)"] --> B["2. Capture Photo\n(Guided framing tips)"]
    B --> C["3. On-Device Preprocessing\n(224x224 RGB + normalize)"]
    C --> D["4. Offline TFLite Inference\n(MobileNetV3, float16)"]
    D --> E["5. Instant Local Result\n(Class, Severity, SQLite)"]
    E -. "Optional when online" .-> F["6. Cloud Sync & Gemini RAG\n(in development)"]

1. Crop Selection: The user selects Coffee (leaf underside), Cacao (pod), or Banana (leaf) prior to scanning so a dedicated specialist model runs for that crop.
2. Offline On-Device Inference: The captured photo is resized to 224x224, decoded to raw RGB, normalized with exact training statistics, and evaluated locally with react-native-fast-tflite. On-device latency has not been benchmarked yet (Phase 8).
3. Actionable Triage & Thresholding: If confidence falls below the calibrated validation threshold, the app instructs the user to retake the photo. Otherwise, it reports predicted class, severity level, confidence, and recommended triage action (healthy, watch, or act now).
4. Local Persistence & Optional Sync: Scans, farms, and plots are stored locally in expo-sqlite with client-generated UUIDs. Optional accounts and idempotent sync through the FastAPI backend (api/) are in progress (Phase 6).
5. Grounded RAG Advisory (planned, Phase 7): When online, the backend will retrieve passages from official technical manuals (ICAFE for coffee, CATIE for cacao, CORBANA for banana) and use Gemini to explain next steps, citing the source and declining when the manuals do not cover the case.

One fine-tuned MobileNetV3-Large model is trained, evaluated, and exported per crop (224x224 RGB input, float16 TFLite, 6.06 MB each). Metrics are on held-out test splits of public datasets; field validation in Costa Rica is still pending (Phase 8):

Every run, including baselines and failed attempts, is logged in docs/EXPERIMENTS.md.

Tensor preprocessing parity between Python (ml/) and React Native (app/src/ml/) is enforced by shared fixture tests.

healthycropping/
├── app/       # Expo React Native (Android, pnpm): on-device TFLite, SQLite, UI
├── api/       # Python 3.12 (uv): FastAPI, Firebase Admin, Cloudinary, RAG
├── ml/        # Python 3.12 (uv): data preprocessing, training, eval, TFLite export
└── docs/      # ARCHITECTURE, PLAN, MODEL_CARD, DATASETS, EXPERIMENTS, DEVLOG

• Mobile: Node.js 20+, pnpm (configured with node-linker=hoisted), Android SDK.
• Backend & ML: Python 3.12, uv package manager.

cd app
pnpm install
pnpm typecheck
pnpm lint
pnpm test
pnpm start

cd api
uv sync --all-extras
uv run ruff check .
uv run mypy src tests
uv run pytest
uv run uvicorn src.main:app --reload

cd ml
uv sync --all-extras
uv run ruff check .
uv run mypy src scripts tests
uv run pytest

Contributions are welcome. Please check our open issues and guidelines before opening a pull request:

• Roadmap & Tasks: https://github.com/keylorpineda/healthycropping/issues
• Community Spotlight: GitHub Discussions Announcement
• Contributing Guidelines: CONTRIBUTING.md
• Architecture Overview: docs/ARCHITECTURE.md
• Code of Conduct: CODE_OF_CONDUCT.md

1. Decision Support Only: Visual symptoms cannot distinguish every microscopic pathogen strain or nutrient deficiency. Laboratory diagnostics remain necessary for high-impact interventions.
2. Field Lighting Sensitivity: Direct specular glare on wet foliage or extreme motion blur can reduce model confidence.
3. Backend Cold Starts: Once deployed, the optional FastAPI sync API is planned to run on Render's free tier, which sleeps after 15 minutes of inactivity. Disease diagnosis and SQLite storage run fully offline on the phone and do not depend on it.
4. Public Training Data: Models are trained on public datasets whose cameras, cultivars and backgrounds differ from Costa Rican farms; real-field accuracy is unmeasured until field validation (Phase 8).

The MIT license in LICENSE applies solely to the source code in this repository. Training datasets remain under their respective upstream licenses documented in docs/DATASETS.md:

• RoCoLe (Coffee — Mendeley Data, DOI 10.17632/c5yvn32dzg.2)
• BRACOL (Coffee — Mendeley Data, DOI 10.17632/yy2k5y8mxg.1)
• CocoaMoniliaDataSet (Cacao — Zenodo, DOI 10.5281/zenodo.17156051, CC BY 4.0)
• UNAB Cocoa Diseases (Cacao — Kaggle serranosebas/enfermedades-cacao-yolov4)
• Banana Leaves Imagery (Banana — NM-AIST DSpace 20.500.12479/3003; DOI 10.1038/s41597-025-04456-4)
• BananaLSD (Banana — Mendeley Data, DOI 10.17632/9tb7k297ff.1)

HealthyCropping (com.keylorpineda.healthycropping) is strictly non-commercial. Starting with v1.0.0, it will be distributed free of charge as an installable Android .apk exclusively through GitHub Releases, with no advertisements or paid subscriptions. Current releases only include the exported .tflite models.

@software{pineda2026healthycropping,
  author       = {Pineda, Keylor},
  title        = {{HealthyCropping: Offline-First Crop Disease Detection for Coffee, Cacao, and Banana}},
  year         = {2026},
  url          = {https://github.com/keylorpineda/healthycropping},
  version      = {0.1.0},
  license      = {MIT}
}

• Keylor Pineda — @keylorpineda
• Jefferson Abarca — @JAbarca74