An irrigation system that waters only when the soil asks for it

The core functionality of the system relies on a humidity sensor that detects the soil’s moisture level underground, triggering irrigation only when the moisture falls below the normal standard. This key design decision not only prevents over-irrigation but also achieves significant water conservation. The project applies hands-on engineering knowledge — Arduino programming and simulation, construction of the main circuits, and repeated testing to optimise performance.

I chose this topic for a combination of practical, environmental and personal reasons. It offers convenience for people who cannot tend to plants during long vacations, it solves the difficulty the elderly face in manually watering plants — a motivation driven by my grandmother’s struggle with carrying heavy water containers — and it promotes environmental protection through water saving, while giving me the hands-on engineering experience I wanted.

I fixed the control parameters and the success metrics before assembly, so that “it works” would mean something measurable. The system triggers at a 30 % soil-moisture threshold and delivers a single watering volume of 1,000 mL. I then ran three rounds of testing at 20 %, 30 % and 50 % moisture to optimise the control parameters, and kept the material cost at about 80 yuan.

The presentation followed the same path as the build, so the audience could see why each decision was made: the problem and the people it serves, the sensing and control hardware, the threshold and volume parameters, and the test rounds that tuned them.

The finished prototype reads soil moisture, compares it with the threshold and waters automatically when the soil is dry. The schematic, the control program and the sensor calibration are documented below.

Looking ahead, this system is expected to create positive impacts on the community and on the elderly group. I plan to make promotion videos introducing the technology to my community, where many residents grow vegetables in first-floor gardens, to help the community save water and avoid over-irrigation. For the elderly, it automates plant watering and eliminates the need to carry heavy water containers, making plant care much easier.

I have already demonstrated the working device to senior gardening enthusiasts in the community and collected feedback on how it behaves in real use, through three interviews covering operational experience and usage needs.

I faced real difficulty with the sensitivity of the humidity sensor, which prevented me from getting the moisture level steady and detectable for a long time. After investigating several means of improving that sensitivity, I finally tried a capacitance sensor — more expensive, but far better suited to my Arduino system. I also adjusted my programming according to advice from my mentor, and after several tests the irrigation system began to function better.

What I take from the experience is that the measuring part of a system deserves the most attention: I had planned the project around water and pumps, and the component that decided whether it worked at all was the sensor. Testing repeatedly, at several moisture levels, is what turned an idea into something that behaves reliably.