iAgro Pest & Disease Fuzzy Risk Model (PDM)

Agrobit S.r.l. — SmartCherry sub-innovation project (SIP10, code RZTHQ), OpenAgri Open Call, Horizon Europe Grant Agreement No. 101134083.

Open-source, multi-crop fuzzy-logic engine that estimates daily pest and disease infection risk from weather data and crop phenology.

What it does

For every day of the season and every configured crop–pest pair, the model computes a 0–100 infection-risk score and a risk class (Low, Moderate, High, Critical), turning weather and agronomic knowledge into actionable, site-specific guidance on when monitoring or treatment is warranted. It supports 18 fruit, nut and vine crops.

Pipeline: weather + crop–pest rule base + pest biological parameters → feature engineering (GDD accumulation, leaf-wetness estimation, moving averages, streaks) → fuzzy Mamdani inference and defuzzification → daily risk CSV + PDF report.

Install

python -m venv venv && source venv/bin/activate
pip install -r requirements.txt

Usage

python main.py                 # interactive crop selection
python main.py --crop Cherry   # single crop
python main.py --crop Cherry Grape   # multiple crops
python main.py --all           # all crops
python main.py --no-report     # skip PDF generation
python main.py --validate      # print validation statistics

Outputs are written to output/: risk_results.csv (20-column daily risk table) and pest_risk_report.pdf (human-readable report).

Files

• main.py — CLI entry point (crop selection, run, validation summary, output)
• config.py — all user-editable settings (paths, fuzzy parameters, risk thresholds, GDD/phenology, leaf-wetness tables, report/colour settings)
• risk_model.py — core engine: data loading, feature computation, membership functions, per-day Mamdani inference (run_model)
• chart.py — build_report(), the multi-section PDF risk report
• test_model_v4.py — tests (parsing, membership functions, end-to-end scoring)
• db_crops_PDM.xlsx — crop–pest fuzzy rule base (crop, pest, humidity, temp, rainfall, risk, type)
• pest.json — per-crop, per-pest biological/phenological parameters
• weather_data.xlsx — daily weather input (date, temp_max, temp_min, humidity, rainfall)
• docs/ — model documentation
• output/ — generated results (git-ignored)

Input / Output

• Input: weather XLSX (date, temp_max, temp_min, humidity, rainfall, auto date format), the rule base XLSX and the pest JSON.
• Output: risk_results.csv (daily per-crop, per-pest scores and classes) and pest_risk_report.pdf.

All tuning parameters live in config.py.

Requirements

• Python 3.10+
• pandas, numpy, openpyxl, xlrd, matplotlib, reportlab

License

Apache License 2.0 — see LICENSE. Copyright 2026 Agrobit S.r.l.

Developed within the SmartCherry sub-innovation project (SIP10, code RZTHQ) funded by the OpenAgri Open Call under the European Union's Horizon Europe research and innovation programme, Grant Agreement No. 101134083.