Investigating the Effect Grow Lights Growth Vegetables Using an IoT-Enabled Vertical Farming System

Beh, Brandon (2026) Investigating the Effect Grow Lights Growth Vegetables Using an IoT-Enabled Vertical Farming System. Final Year Project (Bachelor), Tunku Abdul Rahman University of Management and Technology.

Text BEH BRANDON_ Full Text.pdf Restricted to Registered users only Download (4MB)

Abstract

This study investigates the effects light spectrum on vegetable growth within an Internet of Things (IoT)-enabled vertical farming system, aiming to optimize sustainable urban agriculture. Through controlled experiments on Sawi Hijau and supplementary lettuce datasets, environmental and growth data were systematically collected, monitored, and analyzed using IoT sensors and cloud-based platforms. Machine learning models, including Decision Tree, Random Forest, SVR, and XGBoost, were applied to predict plant growth under varying light and environmental conditions. Among these, XGBoost emerged as the best-performing model, consistently achieving high predictive accuracy across plant height, leaf length, and leaf width. Validation using the "Predict with Sliders" feature demonstrated that predicted values were nearly identical to actual growth observations, with errors typically below 3 percent. Results confirmed that light spectra and intensity play significant roles in determining growth performance, with pink LED and 4000K LED yielding superior outcomes at mid and late growth stages, while sunlight produced consistent average results. This research contributes a scalable framework for IoT-driven predictive farming, offering insights for optimizing plant growth, improving decision-making, and supporting sustainable food production aligned with the United Nation Sustainable Development Goals (UN SDG) 2 and 3