UniTriRob: a robust machine learning regression model for predicting lettuce yields in aeroponic vertical farming

Area of Science:

• Agricultural Science
• Machine Learning
• Environmental Monitoring

Background:

• Aeroponic vertical tower farming offers sustainable food production for Lactuca Sativa (lettuce).
• Accurate biomass prediction is hindered by complex, non-linear relationships between environmental factors and lettuce growth.
• Existing forecasting methods struggle with outliers and heteroskedastic errors in growth parameters.

Purpose of the Study:

• To develop a robust machine learning model for accurate biomass prediction in aeroponic lettuce farming.
• To address the challenges posed by non-linear environmental influences and data variability.
• To enhance yield forecast accuracy and optimize production efficiency.

Main Methods:

• Development of the UniTriRob regression model, a novel machine learning approach.
• Focus on mitigating outliers and heteroskedastic errors in key growth parameters.
• Utilizing parameters such as pH, total dissolved solids (TDS), temperature, electrical conductivity (EC), turbidity, humidity, and light intensity.

Main Results:

• The UniTriRob model achieved a high R-squared value of 97.8386%.
• The model demonstrated a minimized error rate of 0.46%.
• Performance significantly outperformed conventional forecasting methods in experimental validation.

Conclusions:

• The UniTriRob model provides a viable solution for accurate biomass prediction in aeroponic lettuce cultivation.
• This advancement contributes to maximizing production efficiency and yield forecast accuracy.
• The model supports the advancement of sustainable agricultural practices through improved crop management.