Smart Cyber-Physical Systems for Controlled-Environment Agriculture

Funded by UGA President’s Interdisciplinary Seed Grant ($140K, 4/2017-10/2018)

Background

Over the next 30 years, the world’s population will grow by up to 34 percent and urbanization will increase by up to 20 percent. In order to feed this wealthier and larger population, food production must increase by an estimated 70 percent (Food and Agriculture Organization of the United Nations, 2009). Achieving food security by minimizing fluctuations in supply and adjusting to the growth in food demand presents many challenges that will require a major adjustment in current agricultural practices (Magnin, 2016). Precision Agriculture and Food Security can be achieved through the integration of cyber-physical technologies with agricultural and food systems.

 

 

Controlled-environment agriculture (CEA) is a technology-based approach toward food production. The aim of CEA is to provide protection and maintain optimal growing conditions throughout the development of the crop. Production takes place within an enclosed growing structure such as a greenhouse or building. Plants are often grown using hydroponic methods in order to supply the proper amounts of water and nutrients to the root zone. Because of this, CEA does not depend on arable land and food can be produced in or near major population centers. CEA optimizes the use of resources such as water, energy, space, capital and labor. CEA technologies include hydroponics, aquaculture, and aquaponics.

A February 2011 article in Science Illustrated states, “In commercial agriculture, CEA can increase efficiency, reduce pests and diseases, and save resources. … Replicating a conventional farm with computers and LED lights is expensive but proves cost-efficient in the long run by producing up to 20 times as much high-end, pesticide-free produce as a similar-size plot of soil. Fourteen thousand square feet of closely monitored plants produce 15 million seedlings annually at the solar-powered factory. Such factories will be necessary to meet urban China’s rising demand for quality fruits and vegetables.”

 

Figure 1 Smart Cyber-Physical Systems for Controlled-Environment Agriculture

However, the industry’s current practices require considerable energy to power artificial lighting to maintain plant growth on overcast days to meet production schedules. Electricity for lighting can make up 20 – 30% of the production costs in greenhouses and 50-60% in plant factories. Thus, food security and energy & environment sustainability are crucial intersecting grand challenges to be addressed by our society.

Research Goals

This research’s goal is to develop Smart Cyber-Physical Systems for Controlled-Environment Agriculture, to optimize the food production quality, quantity and schedule in balance with operational cost and resource usages. The features of such a system include, but not limited to, the follows:

  1. sense and control environmental parameters (such as temperature, humidity, carbon dioxide, light) and soil properties (such as moisture, nutrient concentration and acidity);
  2. monitor the crop health and provide expert advice and decision support for future actions;
  3. minimize the energy cost of lighting, HVAC and irrigation systems with information from electricity market pricing signals, renewable energy supply, and weather forecasts;
  4. implement the potential correlated benefits of including solar photovoltaic panels on greenhouse structures in terms of energy production, internal shading benefits for both cooling and heating seasons;
  5. optimize food production schedules by connecting with food supply chain and logistics;
  6. automate the seeding, phenotyping and harvesting processes for certain crops with robotics and other intelligent tools;
  7. provide recommendation on optimized greenhouse design and operations based on local weather conditions, food needs, and energy supply profiles.

Descriptions

As a premier land-granted and flagship university, UGA has a large agriculture program across the state of Georgia and many greenhouses for research and experiments. In this project, we will install sensors to measure soil water content, light levels, air and crop temperature, relative humidity, and CO2 levels in greenhouses, and environmental (e.g., temperature, humidity and lighting) control systems, that will allow to optimize the crop growth rate. Cameras will be installed to monitor crop responses to different growing conditions. Additionally, new sensors like net radiometer, infrared radiometer, spectroradiometer, and NDVI sensors may be installed for advanced research. A weather station may also be installed on the roof of some greenhouse for forecast study and comparisons. A small plant factory is also available in the Horticultural Physiology Lab. The plant factory has 18 separate crop production areas with LED lighting and temperature and humidity control.

Figure 2 UGA greenhouses

We will also install renewable energy sources (such as solar panels and wind turbines etc) on greenhouses to form a smart Microgrid with the goal of achieving Net-Zero-Cost (NZC). The renewable energy sources (such as solar panel arrays) serve the primary energy source; while the energy storage and main grid will be used as the secondary. The Microgrid will monitor, predict and control the load, demand and storage based on the weather conditions, crop growth needs and electricity pricing signals. For scalability and resiliency, the decentralized energy management, optimization and control methods will be developed, without relying on a centralized control center. The ultimate goal of the greenhouse Microgrid would be the standalone autonomous operation with minimal energy cost and result the sustainable crop growth, energy and environment.

Figure 3 Illustration of Microgrid Architecture and Elements

We develop a decentralized method for Microgrid that controls their output frequencies, power generations, and voltages during grid-connected, islanding, and synchronization modes. The proposed control system utilizes droop control to quickly balance generation and demand after sudden disturbances. It utilizes distributed computing and control to stabilize the system frequency and voltage, and also to distribute the generation among DGs. Furthermore, the proposed control system adjusts the frequency and voltage to reconnect the microgrid to the main grid by utilizing only local measurements from the neighboring DGs.

 

Faculty

  • WenZhan Song
  • Marc van Iersel
  • Richard Watson
  • Javad Mohammadpour
  • Thomas Lawrence
  • Maric Boudreau
  • Mark Haidekker
  • Yi Hong

Industry Partner