The Self-Watering Garden

Every ATL has plants that die over the holidays. It is not a mystery. The term ends. Everyone goes home. The plants, which did not go home, slowly dry out. By the time the lab reopens — sometimes two weeks later, sometimes more — something is dead that was not dead before. Before every long break, someone always volunteers. "I'll come in and water them." Sometimes they do. Sometimes something comes up. Plants do not send reminders. The solution that actually works is not a more reliable volunteer. It is a machine. A machine that does not go home, does not forget, does not have something come up. A machine that checks whether the soil is dry, decides whether to water, and acts — every few hours, indefinitely, without being asked. This is not complicated. A soil moisture sensor is a probe that measures how well the soil conducts electricity — wet soil conducts better than dry soil, so the reading is higher when the plant needs less and lower when it needs more. That reading can be compared to a threshold. When it drops below the threshold, a small pump turns on. When the reading rises again — because the soil is now wet — the pump turns off. Sensor reads. Decision is made. Pump acts. Sensor reads again. This is a closed loop. The output — water delivered — feeds back into the input — soil moisture — and the system corrects itself without anyone touching it. Today, your team is going to build one.