Design and Research of TreatMate: A Smart Public Dog Food Vending Machine

Abstract

Urban regions face persistent challenges related to irregular stray-dog feeding, unmanaged plastic waste accumulation, and inconsistent access to controlled animal-food dispensing. This paper presents Treatmate, an IoT-enabled public dog food vending machine that integrates verified digital transactions with a recycling-oriented activation pathway to ensure controlled and incentivized food dispensing. The system is built around an ESP32 with an OV2640 camera module, while a Tiny CNN (96×96 input, quantized int8, TensorFlow Lite Micro) performs on-device bottle material classification to confirm whether the item inserted for recycling is plastic before rewarding the user. The dispenser uses a dual-actuation mechanism with a DC geared motor and servo motor, and real-time stock monitoring is provided through a load cell interfaced via an HX711 amplifier. A cloud-based backend developed using Node.js manages secure transaction validation, device synchronization, and centralized logging over HTTPS communication. In addition to automated dispensing, the system records recycling activity and feeding patterns within a centralized database. A dedicated analytics layer processes historical dispensing and bottle-insertion records to identify feeding trends and correlate food distribution with regional stray-dog demand. The proposed system demonstrates a scalable, sustainability-oriented smart infrastructure model integrating embedded vision, cloud computing, recycling incentives, and data riven urban welfare management.