Problem Statement

Plant automation systems such as the one we designed have many applications, both on a consumer and industrial scale. Our main purpose in designing this system was to give Mike Dyer a software platform through which he could automate and control the watering of a small set of plants, and also get useful, long-term feedback on the health of the plants. We wanted the system to be robust, energy and cost efficient, and user friendly. In addition to aiding the operators of the greenhouse, we also wanted to create a system that would serve as a proof of concept for larger scale IoT based plant control systems.

The idea of automating plant watering is not a new one. Many people have attempted to create DIY plant watering systems using Arduino, Raspberry Pi, and other similar microcontroller platforms. These systems usually involve a moisture sensor feeding readings back to the controller, and a simple actuation system. They are mostly hobbyist-level projects, designed for use in households. With regard to commercially available solutions, there exist few similar products. These usually come in the form of “kits”, which can be installed into a plant or set of plants, and cost around $50-$200. Some of these products are simply cleverly designed pots and reservoirs, which are able to distribute the water supply over a period of time. Others are timer-based irrigations systems, which periodically water the plants based on a timer. In this sense, not many commercially available products even attempt to solve the problem we are attempting to solve.

History of Jeanette Goldfarb Plant Growth Center

The Jeannette Goldfarb Plant Growth Facility was constructed during a $3.5 million physical expansion of Washington University’s Department of Biology. The purpose of the facility is to study experimental plant growth and research. Currently the Jeannette Goldfarb Plant Growth Center has no automated mechanism to monitor and maintain the health of plants. The facility is equipped with 24/7 lighting and a temperature regulation system. Watering is done manually for the most part with a few exceptions based exclusively on manual programming and timers. Michael Dyer as no method by which he is able to water his plants remotely and make sure that they do not die other than sending a student in to check.

Objectives

  • The main objective of this project is to provide our client, Michael Dyer, with an IoT system to automate the operations of watering an array of plans in the greenhouse.
    • We will be relying on Michael’s knowledge of plant growth and of the Jeanette Goldfarb Plant Growth Facility’s hydraulic systems.
    • We will need Michael’s advice on optimal scheduling for watering our plants
    • We will use Michael’s equipment catalog to locate future equipment such as solenoid valves, peristaltic pumps, higher quality soil sensors, etc…    
  • The end-user of our system should be able to monitor the status of plants’ health 24/7 from anywhere in the world on a computer or mobile device.
    • We will be working through particle’s cloud database
    • Collect data for display through our custom UI
  • User will be able to control select environmental inputs and view both historical and real time data for individual plants through a simple and easy to use GUI.
    • HTML/Javascript/CSS stack
    • Potentially hosted on AWS EC2 instance for accessibility.
  • We would like to verify different ‘optimized’ schedules for our system to autonomously take care of plants.
    • Actuate light, water, temperature autonomously
    • Inform algorithms with industry best practices for growing of specific plants