When Water Keeps Flowing: A Low-Cost ESP32 System for Preventing Household Water Waste

<jats:p>This paper presents Each Drop Counts, a low-cost, ESP32-based domestic water-conservation prototype designed to allocate and safeguard a fixed water quota for each flat or household member, for example, a 500-liter daily allowance per flat, resetting automatically at 00:00 local time. Beyond simple quota tracking, the controller distinguishes ordinary, intermittent household water use from a continuous, unattended flow event. When water flows without interruption past 50 liters, the system issues a buzzer warning and closes a servo-actuated 3D-printed valve for 30 seconds; if the flow resumes and continues uninterrupted for a further 10 liters, a second warning and closure follow; and if flow persists for an additional 3 liters, the valve locks permanently, requiring an authorized Bluetooth command to reopen. This staged escalation was designed to mirror the behavior of commercial tank-filling systems that shut off automatically once filling completes, while still giving the user two clear opportunities to intervene before restriction. An emergency Bluetooth override allows an authorized user to claim an additional allowance, nominally around 200 liters, for genuine high-need situations, with this limit configurable per user requirement. A pay-to-recharge mechanism, allowing a user to top up an exhausted quota, was also considered as part of the system's intended design and is discussed as a configurable extension, though it remains unimplemented in the current firmware. A 16×2 LCD provides local feedback on accumulated volume and system state, while the flow sensor's placement upstream of the valve reduces the chance of downstream drain-down being misread as post-closure consumption. The prototype demonstrates working volume display and valve-closing behavior on hardware, built around efficient, near-complete utilization of the ESP32's onboard GPIO, I2C, PWM, and Bluetooth capabilities. The paper documents the control logic, hardware architecture, test evidence, limitations, and a roadmap toward IoT, Android, and edge-AI integration.</jats:p>