Optimizing Indoor Farming: Deep Learning for Predicting Plant Growth under LED Light Treatments

Tanvi Verma

Abstract:

Indoor farming has emerged as a rapidly growing
industry that harnesses controlled environmental conditions to
cultivate crops. A key component of indoor farming is the
utilization of LED lighting, which serves as the primary source
of light for plant growth. In this study, our objective is to
optimize indoor farming practices through the application of
deep learning techniques, specifically by predicting the growth
of plants under different LED light treatments in controlled
agricultural environments. To achieve this goal, we employed
existing machine learning methods and proposed a novel deep
learning approach that incorporates the effects of LED light
spectrum on plant growth to estimate plant biomass. Our deep
learning model, BioNet, utilizes 1D convolutional neural network
(CNN) to extract spatial features from the light spectrum data.
Through extensive experimentation and analysis, we demonstrate
that our deep learning method outperforms other conventional
methods, showcasing its potential to enhance our understanding
of the impact of LED light on plant growth.
Our research provides valuable insights into optimizing indoor
farming by uncovering the relationship between LED light
treatments and plant biomass. BioNet serves as a valuable
tool for farmers, enabling informed decisions on LED light
selection, leading to improved efficiency and productivity in
indoor farming. This study contributes to advancing indoor
farming techniques through deep learning, opening new avenues
for exploration and highlighting potential improvements in the
field.