A model for statistical forecasting of menu item demand

Area of Science:

• Operations Research
• Healthcare Management
• Food Science

Background:

• Foodservice planning relies heavily on accurate demand forecasting for operational efficiency.
• Current hospital food management systems often struggle with tedious and inaccurate menu item demand forecasting.
• Effective forecasting is crucial for production, workforce, facility, and resource allocation decisions.

Purpose of the Study:

• To develop and evaluate a more accurate and efficient method for forecasting menu item demand in hospital foodservice.
• To demonstrate how statistical time series predictions can improve demand forecasting accuracy.
• To reduce the labor and inaccuracies associated with traditional forecasting methods.

Main Methods:

• Utilizing statistical time series models to forecast meal tray counts.
• Multiplying meal tray count forecasts by average menu item preference percentages to predict demand.
• Developing simple worksheets for manual forecast generation and data collection.
• Implementing a system for ongoing data collection, including tray counts and forecast errors.

Main Results:

• The proposed system offers potential for significant cost reductions at various levels of foodservice operations.
• Accurate demand forecasting minimizes the need for costly plan adjustments.
• The method allows for generating more precise menu item predictions with minimal labor increases.

Conclusions:

• The developed forecasting system provides a more accurate and less labor-intensive approach to menu item demand prediction in hospitals.
• Statistical modeling and data collection by-products can transform foodservice planning from a chore into an efficient process.
• Further research is ongoing to quantify the cost-benefit relationship and expected savings more explicitly.