Automatic nutrient estimator: distributing nutrient solution in hydroponic plants based on plant growth

Abstract

Background The primary objective is to address the specific needs of plants at different growth stages by delivering precise nutrient concentrations tailored to their developmental requirements. Challenges such as uneven nutrient distribution, fluctuations in pH and electrical conductivity, and inadequate nutrient delivery pose potential hindrances to achieving optimal plant health and yield in hydroponic systems. By overcoming these challenges, the hydroponic farming community aims to enhance the accuracy of nutrient dosing, streamline automation processes, and minimize resource wastage. Hydroponics, a cultivation technique without soil, facilitates the growth of organic vegetation while concurrently minimizing water use and eliminating the necessity for pesticides. In order to achieve effective cultivation of hydroponic plants, it is essential to maintain a controlled environment that encompasses essential factors such as temperature, carbon dioxide (CO2) levels, oxygen availability, and appropriate lighting conditions. Additionally, it is crucial to ensure the provision of vital nutrients to maximize output and productivity. Due to the demanding nature of a hydroponic farmer’s schedule, it is necessary to minimize the amount of time dedicated to nutrient management, as well as pH and EC adjustments. Methods In order to determine and deliver the proper amount of vital nutrients, such as nitrogen, phosphorus, and potassium, based on the plant growth stage, we presented an automatic hydroponic nutrient estimator in this system. We noticed that the plant’s nutrient consumption varies depending on its stage of growth according to plant psychology. Four peristaltic pumps with the necessary sensors are controlled by an Arduino board in the suggested system. Both filling and draining the water are done using each pump. To identify the plant stage, we apply the Plant Growth Stage Identification algorithm to encompass the seedling, vegetative, flowers, and fruit stages. Results The experimental results reveal that the Growth Stage Identification algorithm obtains 97.5% accuracy for the first 5 weeks with 1,715 ppm of nutrition ingestion, identifying the vegetative state. The flowering stage is identified with 97.5% accuracy in the 6–9th week with 2,380 ppm of nutrition consumption, and the fruiting location is determined with 99.4% accuracy in the last 10–15th week with 2,730 ppm of nutrition consumption.

Keywords

• Automatic hydroponic
• Nutrient estimator
• Mask R-CNN
• Arduino UNO
• IoT sensors
• Micro nutrients