Follow The Light

Buying food can be very expensive. And it can be even more expensive to buy healthy food. One solution to reducing these costs, would be growing your own healthy food at home. However, many people do not have access to a garden space or green house where they can grow food all year long. In addition, many people may not have a window that faces the sun or that receives optimal sunlight for plants to grow indoors. My goal was to create an efficient way to grow plants inexpensively in the home.

My idea for an indoor plant growing system was a solar tracker; a machine that could be placed beside a window and would follow the sun to achieve optimum light levels. I used photoresistors (light sensors), an Arduino microcontroller (programmable mini computer) and servo motors to detect, follow and maximize the light a plant receives. I also wanted to use LEDs to achieve optimum light levels for maximum growth on cloudy days or if the only available window does not receive continuous direct sunlight, and to come on before sunrise and after sunset to achieve the best photoperiod for plant growth.

As part of my design, I had to do background research to figure out the ideal LED colour mixture (2) and photoperiod (1) for an example plant (spinach). Since objects reflect their own color, I knew I needed light with little to no green on the RGB light scale(5). After reviewing several scientific journals (4) from the field of agriculture, I found out that the best color for spinach is a mix of 65% Red, 25% Blue, and 10% white (3).

The first step in the design was creating the solar tracker. I used 4 light sensors facing different directions. I wired them to the Arduino and motors. I then created a code that measured the differences and told the motors to move in response. After this I made a separate code and wiring for the LEDs. I coded it so that when light level get below a certain threshold, the LEDs come on.

I tested the light levels, differences, and angles of the motors over several days to test and refine the margins and thresholds in the code for the LEDs and motors. I also refined the design and placement of the sensors to account for ambient light.

I was able to collect 4 days of data. There were some sunny days and some overcast days. I measured the light levels horizontally and vertically. I graphed the horizontal light levels (right and left) (Fig 1.0), the differences between right and left (Fig 1.1) and the angle of the plant stand (Fig 1.2). Shown above are the results for a sunny day. I also included the horizontal light levels on an overcast day for comparison (Fig 1.3).

The sun moves from east to west across the sky. I expected the light level to increase at sunrise and decrease as the afternoon goes on. I expected the photoresistor facing east to be reading higher light levels than the west photoresistor in the morning, and then for them to switch around noon. On the sunny day the results followed the hypothesis, but on the overcast day (Fig 1.3) there was too much ambient light and no direct sunlight. On the overcast day the expected change over from east to west, switching at noon, was not observed.

The differences across right to left light levels were greater in the morning and evening because the sunlight was more to one side during that time.  At mid-day, the differences were less because the photoresistors were getting very similar sunlight.

For the vertical light levels, the margin didn’t change much and didn’t affect the angle of the plant stand. The plant was always angled the same way, upwards towards the light.

The overall goal was for the horizontal angle of the plant stand to follow the sun over the course of a day. It did on the sunny day.

The next thing I tested was the LEDs. Their purpose was to come on when the light was low to compensate. They came on but they would turn off the next time the code looped. This was because when the light came on, it detected the LED light, so it turned it off. Then the cycle repeated. I fixed this by editing the code so when it detects the light is low. It turns it on for 3 hours then it can't come on for another 15 hours.

Overall on a sunny day, the machine followed the sunlight across the sky, and this was the original goal. But they're were many stuggles, such as to measure the biggest difference across the horizontal photoresistors, I had to follow engineering principles and iterate the design to achieve a measurable difference. This could be because the photoresistors are very sensitive. The other possible explanation is that there is minimal light difference overall and that following the light wouldn't be an advantage in growing plants.

LEDs are widely used in indoor plant growing systems. This addition to my design was meant to increase and improve the photoperiod and plant growth, and it was successful.

One thing I would like to do going forward is to begin testing with an actual plant. I think it would be interesting to compare my plant to a plant growing on a windowsill normally. I would also add more adaptations to the code, like making it adapt to the season or weather. I could also improve the LEDs by making them compensate based on the light level. I also would like to do a little more research to the light photoperiod and colour mix and maybe even do an experiment on it.

I would like to thank my parents for helping me with this this project. They supported me throughout and helped me do things like source vintage meccano and remind me to get up at sunrise to start my data collecting. I thank them for that.

1. General Photobiology & PAR Overview Photosynthetically Active Radiation (PAR) and Plant Response. (Encyclopedia / Review material summarizing how wavelengths between 400–700 nm affect

photosynthesis; foundational context rather than a primary experiment)

2. Research Article on Wavelength Effects in Spinach Impact of Light Wavelengths on Photosynthetic Rates in Spinach.

3. Yorio, N.C. et al. Improving Spinach, Radish, and Lettuce Growth Under Red LEDs With Blue Supplementation. HortScience. 2001, 36(7), 380–383.

4. Zeevaart, J.A.D. & Talon, M. Effects of Photoperiod on Growth Rate and Endogenous Gibberellins in the Long-Day Rosette Plant Spinach. Plant Physiology. 1971, 47(6), 821–827.

5. Zou, Z. et al.Optimization of Artificial Light for Spinach Growth in Plant Factory Based on Orthogonal Test. Plants (Basel). 2020, 9(4), 490. DOI: 10.3390/plants9040490