This project presents an adaptive paper towel dispenser system designed to reduce waste through behavioral learning and user engagement instead of restrictive dispensing controls. The system integrates an ESP32 microcontroller with capacitive proximity sensing, ultrasonic roll monitoring, and an adaptive learning algorithm that continuously adjusts dispense length based on real usage patterns. A Multiple Dispense Ratio control mechanism maintains user satisfaction and encourages conservation through eco-feedback displays. The system operates from a standard 4×D-cell battery pack, consuming less than 2.4mA in sleep mode and initiates dispense cycles within 500ms. Network connectivity enables automated maintenance alerts and centralized monitoring through a Message Queuing Telemetry Transport (MQTT) dashboard. Field data collected from three campus dispensers informed the adaptive algorithm design, targeting 1-5% waste reduction depending on optimization level. The design addresses limitations in current commercial dispensers that use fixed-length dispensing or time delays, shifting the paradigm from user restriction to transparent engagement in resource conservation.
