CafeteriaSense

AI-powered food waste prediction and environmental impact intelligence

Input Data (20 Features)

Enter all 20 factors affecting food waste.

Quick Start: Load a Demo Template

Temporal Factors

Demand Factors

Menu & Preparation

Environmental Factors

Operations

Prediction Results

Fill all 20 fields and click predict

4-Week Forecast

Predicted waste for the next 28 days. Green (low) to Red (critical).
Make a prediction first

Environmental Impact

Real-world impact: CO₂, water, meals wasted, disposal costs.
Make a prediction first

School Benchmark

Compare to national averages from USDA school food waste studies.
Make a prediction first

How The Model Works

Understanding the AI: what drives predictions, how to interpret results, and important limitations.

The Machine Learning Model

What it does: Predicts food waste per student based on 20 operational factors

How it was trained: 50,000 simulated realistic school scenarios with noise (like real data)

Accuracy: Predicts waste within ±0.04 lbs/student on test data (80% R² score)

What it is NOT: A perfect forecast. Real schools have unexpected events, special meals, and variability.

What Drives Food Waste (Feature Importance)

Menu Complexity (Highest Impact) 40-45%

Complex menus (many items, prep steps) lead to more waste. Simple 5-item menus reduce waste.

Student Attendance 25-30%

Lower attendance = more food prepared than eaten. Mondays & Fridays have lower attendance.

Portion Sizes 12-15%

Larger portions = more waste. 9-10 oz portions have 15-20% less waste than 12-14 oz.

Staffing Ratio 8-10%

Understaffed kitchens = more waste. Adequate staffing improves quality control.

Equipment & Other Factors 5-10%

Poor equipment, bad weather, exams, holidays also impact waste.

⚠️ Important Limitations (Responsible AI)

This model is a TOOL TO ASSIST, not a decision maker

Use these predictions to:

  • ✓ Identify high-impact areas for improvement
  • ✓ Track waste trends over time
  • ✓ Plan interventions and test their effects
  • ✓ Communicate with food banks and donors

Do NOT use for:

  • ✗ Punishing or blaming kitchen staff
  • ✗ Making firing/hiring decisions
  • ✗ Sole basis for operational decisions

Known Limitations

❌ Does not account for:

  • Special events (sports events, field trips, fire drills)
  • Cultural food preferences and dietary restrictions
  • Food safety incidents or recalls
  • Seasonal farm-to-table programs
  • Emergency closures or unexpected schedule changes

⚠️ May not be accurate for:

  • Very small schools (<300 students) or very large (>2500)
  • Schools with alternative food models (fully vegetarian, allergen-free, etc.)
  • Schools outside New Jersey or with different demographics
  • Schools that already have waste reduction programs in place

📊 Data Quality: Model trained on realistic but simulated data. Real waste varies due to unmeasured factors (student moods, viral videos, social trends, etc.).

Responsible AI Practices

✓ Transparency

All 20 input factors are visible. You can see exactly what drives predictions.

✓ No Personal Data

Model uses only school-level operational data. No student names, IDs, or dietary restrictions.

✓ Beneficiary-Focused

Recommendations prioritize food rescue (feeding those in need) over just waste reduction.

✓ Explainability

After each prediction, see which factors had biggest impact on that specific result.

✓ Actionable Output

Get specific, feasible interventions—not vague recommendations.

Food Bank Locator

Find nearby food banks to donate excess food.

Research Sources

All data and benchmarks based on peer-reviewed research and government data.
USDA Economic Research Service - Food Loss and Waste
Used for: National food waste benchmarks in school cafeterias (0.225 lbs/student baseline)
→ View ERS Report
EPA WARM Model (Waste Reduction Model)
Used for: CO₂ emissions calculations (3.3 kg CO₂ per pound of food waste)
→ EPA WARM Tool
School Nutrition Association - Annual Survey
Used for: Understanding school food service operations and meal preparation factors
→ SNA Website
Journal of Foodservice Business Research
Used for: Understanding relationships between menu complexity, portion sizes, and waste generation
→ Journal Link
NOAA Climate Data
Used for: Weather patterns and temperature/humidity effects on food waste
→ NOAA Data
Feeding America - Food Bank Network
Used for: Food donation logistics and partnerships with local food banks
→ Feeding America
NJ Department of Agriculture
Used for: New Jersey-specific food waste benchmarks and school food service data
→ NJ Agriculture
Environmental Education Research
Used for: Behavioral factors in food waste and effectiveness of reduction interventions
→ EER Journal

AI Recommendation Strategy Sources

Cornell Food and Brand Lab - Tray Size Study
Used for: Effect of smaller trays on reducing portion sizes and food waste
→ Cornell Food Psychology Lab
USDA Food Waste Reduction Research (2022)
Used for: Student choice programs and their effectiveness in reducing plate waste
→ USDA Research
MIT Sloan Management Review - Predictive Analytics in Food Service
Used for: Using demand forecasting and AI to predict daily food needs
→ MIT Sloan Review
EPA Food Waste Prevention Guidelines
Used for: Composting programs and waste diversion strategies
→ EPA Waste Prevention
Cornell University Food Science Program
Used for: Menu engineering and staff training on food safety and waste reduction
→ Cornell CALS