Last spring my daughter, Sarah, needed a project for the science fair, and since she has always been interested in scientific measurements, we decided to try and figure out when it was time to water her mother’s plants. Since we’ve fielded a lot of calls from customers asking about using Arduinos (user-programmable microprocessors) lately, I thought I would kill two birds with one stone and give one a try. My preference would have been the speed and simplicity of a METER data logger, but I was curious about how practical and cost-effective this method might be for taking measurements.
The Arduino is an inexpensive, user-programmable microprocessor on a circuit board that has exposed analog inputs for measuring voltages and digital ports for measuring incoming digital signals. It can also run displays and is programmed by an Arduino IDE running on your computer.
I purchased a book called Arduino Recipes that taught us the basics of Arduino programming, which was pretty straightforward. The Arduino board itself has rows of pinheaders, so I brought some of the male pinheaders from work and soldered all the wires to them, in preparation to attach the water content sensor. It looked medusa-like with all the wires coming off the pinheaders, but we could then just hook up kid-friendly snap circuits and try some elementary tests to get used to the system.
We hooked up Decagon’s (now METER) analog water content sensor (EC5) first and started measuring. It has a really nice calibration equation supplied by METER, so we used that for a while to measure water content. We took one of mom’s dry plants and measured before and after watering and used the readings to make a linear relationship between the reading on the sensor when it was dry and the reading on the sensor when it was wet.
Sarah was excited to get the digital soil moisture sensor integrated because we could then measure water content AND electrical conductivity (EC) to get an idea of the fertilizer in the soil. We used my work colleague’s code to read the digital sensor output, which worked quite well. It only took a few minutes to insert his piece in the code into our program and start reading water content. Our biggest challenge was that Sarah wanted to display this to mom to make sure she knew when to water the plants. So she and I then had to figure out how to integrate an LCD display. Luckily, all the details were on the Arduino website. We just cut and pasted the code into our program and then did all the wiring.
Finally, we had it all put together, and we inserted the 5TE digital sensor into the pot. It worked, but the device was large and unwieldy. Mom wasn’t happy that we were putting it right in the middle of her clean living room, but Sarah pointed out that we have to make sacrifices for science, so we put the sensors in the soil, set up the display, and ran it for about a week. Sarah took water content data morning and night and watered it when it reached our “dry” point. She took the finished system to the science fair and was excited to find a few future customers.
Are Arduinos practical for use in your experiments?
It depends. Sarah and I found out that it just doesn’t take a lot to integrate a sensor into the Arduino system and be able to make measurements. However, if we were to try the above experiment long-term, the biggest challenge would be all the details in the system. We’d need a circuit board, a power supply, a data logging interface board, and a box to put it in, and if we were going to set it outside, that box would have to be waterproof. We’d also need ways to connect the sensor to the circuitry, and all these things take time and resources. For me, the take-home message was that Arduinos are a lot of fun, and might fit your application exactly the way you want. However, you’ll need time (often a lot of it) to spend making sure it’s waterproof, doing all the programming, writing a code durable enough to fit your field applications, and getting the hardware prepped. In fact, Decagon support staff take calls every week from frustrated do-it-yourselfers who’ve found this is not as easy as it seems. Thus, in my opinion, an EM50 or Campbell Scientific data logger are more practical options than an Arduino-like microprocessor.
Are Arduinos cost effective?
A lot of scientists want to make measurements out in the field with small budgets. I am certainly one of those. Arduinos are $85 versus a complete data logger that costs several hundred dollars. However, people tend to forget that things like labor even cost discrepancies.
So, if you have plenty of time, want the versatility, and you love this stuff, go ahead and make an Arduino sensor, but at the end of the day, the cost shouldn’t be a driver, because there are data loggers that can do the job of an Arduino more simply and quickly, without all the hassle.
Get more information on applied environmental research in our