Exploring the Future of AI in Agriculture: Ziye’s Story

Background

‍

"During my master’s studies at Wageningen University and Research, I was educated as a plant scientist and became deeply attracted and inspired by the promising future of potential applications of AI to greenhouse horticulture. This fascination led me to choose a master’s thesis topic that combined AI with greenhouse horticulture crop growth model. After completing my master’s degree, I returned to China, just as the concept of smart agriculture was rapidly gaining attention from investors. I joined a smart agriculture startup, where I engaged in research and product development focused on a crop-model-oriented decision-making system. This work allowed me to visit various farming environments and interact with a wide range of growers. My hands-on experiences in the field profoundly reshaped my understanding of the challenges of applying data-driven approaches in real-world agricultural production. The numerous issues I encountered sparked a desire to delve deeper into the Sim2Real challenges specifically in the agriculture domain. This became the primary motivation for pursuing a PhD.

‍

During this period, I had the chance to meet Dr. Congcong Sun, an expert in control theory. Our insightful discussions made me realize that the integration of control theory with agriculture is a crucial but underexplored area in the applications of AI to agricultural production. I thought, why not take on this challenge myself? And so, I decided to join this lab."

‍

Working Experience

‍

"In this lab, my research is centered on plant-oriented climate control within controlled environment agriculture (CEA) systems. Specifically, I am tackling the Sim2Real challenge in the AI-Agriculture domain by employing reinforcement learning techniques. My work involves integrating visual signals in crop growth models and investigating the potential of image-based control techniques in CEA, thereby advancing the development of data-driven autonomous CEA control in real world."

‍

Impact

‍

"One of the highlights of my time in these two years is leading a team to participate in the 4th Autonomous Greenhouse Challenge. Many of my colleagues’ research are related to greenhouse production systems. We assembled a team and combined the fruits of our individual research projects, achieving a top five ranking in the first two stages, which earned us the opportunity to test our fully autonomous controller in a real greenhouse compartment. This challenge is ongoing, and in the coming months, I will see my findings and AI-driven approach take control of the greenhouse, ultimately cultivating and producing real tomatoes. All of this aligns perfectly with my original intention of pursuing a PhD.

‍

Looking back at these two years in the lab, I believe it has been the right choice for me. The lab has a long-standing history of research in agricultural control and decision-making thus being able to offer profound insights into the core of these challenges, which is invaluable for effectively applying AI approaches to agriculture. This reinforces my confidence that during my PhD, I will have the opportunity to contribute to the advancement of this field."

‍