In the heart of Pune, India, at the Dr. Vishwanath Karad MIT World Peace University, a groundbreaking study is reshaping the future of agriculture and energy efficiency. Sankalp Kadam, a researcher from the Department of Electrical and Electronics Engineering, is leading a project that combines machine learning and explainable AI to predict the growth of Thai basil in hydroponic systems. This isn’t just about growing herbs; it’s about revolutionizing how we approach sustainable farming and resource management.
Hydroponics, the practice of growing plants in nutrient-rich water rather than soil, has long been hailed as a sustainable solution to modern agricultural challenges. Deep-water culture (DWC), a specific hydroponic technique, takes this a step further by providing plants with a continuous supply of nutrients and oxygenated water. This method has shown promising results in increasing plant growth, but predicting and optimizing this growth has been a complex puzzle.
Kadam’s research, published in the IEEE Access journal, titled “Machine Learning and Explainable AI for Thai Basil Growth Prediction in Hydroponics,” delves into this puzzle. “The goal is to create a system that not only predicts plant growth but also explains why it makes certain predictions,” Kadam explains. This is where explainable AI comes into play. By using models like Shapley additive explanations (SHAPs) and local interpretable model-agnostic explanations (LIMEs), Kadam’s team can provide clear, understandable justifications for their predictions.
The study monitored Thai basil, a medicinal plant, over a one-month period, tracking environmental factors like temperature, humidity, solar radiation, pH, and total dissolved solids (TDS). The team also recorded physiological parameters such as green area and plant height. Six machine learning models were employed to estimate growth, with the best model selected using an ensemble voting regressor method.
So, why does this matter for the energy sector? Sustainable agriculture is intrinsically linked to energy efficiency. Hydroponic systems, with their precise control over environmental factors, can significantly reduce water and nutrient waste. Moreover, the ability to predict plant growth can optimize energy use in greenhouses, reducing heating, cooling, and lighting costs. As Kadam puts it, “The more we understand about plant growth, the better we can manage our resources.”
The implications of this research are vast. As urban populations grow and arable land becomes scarcer, hydroponic farming offers a viable solution. By integrating machine learning and explainable AI, Kadam’s work paves the way for smarter, more efficient farming practices. It’s not just about growing Thai basil; it’s about growing the future of sustainable agriculture.
This research, published in IEEE Access, which translates to “Access to Information and Education in Engineering and Technology,” is a significant step forward. It opens doors to new possibilities in precision agriculture, where data-driven decisions can lead to unprecedented levels of efficiency and sustainability. As we look to the future, Kadam’s work serves as a beacon, guiding us towards a world where technology and nature work hand in hand to create a more sustainable tomorrow.