Amador, Kathleen T. <br/><br/>
Balai:   iot power and water management system with ai for mode-control, recommendations and capping mechanism/ 
Kathleen T. Amador, Adrian Gabriel B. Arcilla, Paul Christian M. Castillo, Margaret Zena A. Dalire, Rein Kirsten E. Solpico, and Gino D. Vidal.-- 
 - Manila:	  Technological University of the Philippines,	  2025. 
 - xii, 95pages:  29cm. 
<br/><br/> Bachelor's thesis<br/> 
<br/><br/>College Of Engineering.--<br/> 
<br/><br/>Includes bibliographic references and index. 
<br/><br/> BalAI: IoT Power and Water Management System with AI for Mode-Control,<br/>Recommendations, and Capping Mechanism introduces a streamlining approach to<br/>household utility regulation within a rapidly growing urban environment and booming<br/>demand for core needs. Most of the traditional household utility infrastructures remain<br/>traditional or non-monitored and rarely have facility to communicate in bidirectional<br/>manner, resulting in inefficient resource utilization and high consumption patterns. A lot of<br/>solutions in smart home technologies work in isolation and do not provide intuitive, smart<br/>control based on the behavior of users. In addition, low cost and access in development<br/>areas still raise major issues. Two separate monitoring devices were built using ESP32<br/>microcontrollers. The power module incorporated sensors such as ZMPT101B for voltage<br/>and ACS712 for current, while the water system used a YF-S201 flow sensor and a<br/>motorized solenoid valve. Both modules were mounted on a custom PCB and supported<br/>by relay control and voltage regulation. A Flutter-Dart Language and Firebase based mobile<br/>application enabled control, monitoring and notifications. Moreover, a PPO (Proximal<br/>Policy Optimization) reinforcement learning model was carried out by using<br/>TensorFlow.js. The system was tested from mid-February to mid-April, learning and<br/>responding with data and feedback in real time. The figures speak for themselves: 8 cubic<br/>meters of water and Php 348.7 worth of electricity were saved using optimized usage caps<br/>and AI-generated advice. It also got better over time, learning from what users did. BalAI<br/>proves that a locally adaptable, AI-integrated solution can empower households to manage<br/>utilities effectively, reduce waste, and support sustainable living through practical, user-<br/>friendly technology. 
<br/><br/><label>Subjects--Topical Terms: </label><br/>IoT utilities<br/>Power monitoring<br/>ESP32 module
<br/><br/><label>LC Class. No.: </label>BTH TK 870  / A43 2025