🏠 Αρχική Σελίδα 📄 CV 🌱 OSDE Tool 📈 Financial Dashboard GitHub

Spyros Papastergiou | phytai.com

🍅

Industrial Tomato Quality Control Pipeline

Python · Pandas · NumPy · Matplotlib · openpyxl · Chart.js · 14 charts · HACCP simulation · 28-30 Brix aseptic paste

PythonPandasNumPy MatplotlibopenpyxlHACCP/ISO 22000 Six Sigma / CpkKyknos S.A. benchmark

View Full Pipeline on GitHub

Industrial Tomato QC Dashboard 28-30 Brix Aseptic Tomato Paste · Synthetic Data · Greek Tomato Season 2025

Loading…

Run the demo with 720 hourly batches, or drop your own tomato_qc_data.csv here

Load your own CSV

From: To: Status: Parameter (out-of-spec): —

Accept Batches

—

Correction

—

Critical Waste

—

Production Yield

—

Accept + Correction

Monthly QC Cost

—

Correction + Waste

HACCP Season Savings

—

×3 months Greek season

Batch Status Distribution —

Hourly samples classified against the spec limits.

Daily Trend — Batches & Cost

Stacked bars show the daily outcomes, the line shows the daily QC cost.

Recent Batches last 20

Date	Time	Batch	BRIX	pH	Acid%	Bostwick	Status	Cost €	Deviations

HACCP / ISO 22000 Impact Simulator

Waste reduction: 25%

Operational cost reduction: 15%

Months/year of production: 3

HACCP investment (€): €35.000

Monthly savings

—

Annual savings

—

Batch value saved

—

already inside the savings

ROI break-even

—

months

Parameter Histograms vs Spec Limits

Each histogram shows one parameter across the 720 samples. Green line: target. Red lines: spec limits (min and max). The more values outside the red lines, the bigger the quality problem.

Pareto — Failure Mode Analysis 80/20 Rule

Which parameters cause 80% of the failures, so the monitoring effort goes there first.

Process Capability (Cp / Cpk)

Six Sigma benchmark: ≥1.33 world-class · ≥1.00 acceptable · <0.67 inadequate.

Hour-of-Day Failure Pattern

Failure cost by hour of day. With real data it shows problem shifts; the synthetic data has no shift effect.

Day-of-Week Pattern

Failure cost by day of the week, to spot recurring patterns in real data.

Parameter Performance Summary

Parameter	Min	Target	Max	Mean	Std	Min observed	Max observed	Out-of-spec count	Cpk

High-Resolution Static Charts Generated by Python pipeline

📊

14 Publication-Quality Charts

Generated by the Python pipeline using Matplotlib & Seaborn:
Pipeline flow · Status distribution · Daily trends · Parameter distributions (×8) · HACCP Before/After · Pareto analysis · Correlation heatmap · SPC Control Charts (±3σ) · Process Capability (Cp/Cpk) · Failure heatmap · Boxplots by status · ROI break-even · Risk Priority Matrix (FMEA)

View all 14 charts + full pipeline on GitHub

Run generate_pipeline.py --regen to regenerate all charts locally

Click Run demo above to load one simulated month

Or load your own tomato_qc_data.csv with the same columns