SmartBin

IoT Waste Monitoring System. Deployed at PT Persita

TypeProfessional · Real Client. PT Persita Tangerang Raya

Year2025

Problem

PROBLEM

PT Persita's stadium had a problem familiar to every large venue: bins overflowing during events, manual inspection cycles that couldn't keep pace with crowd density, and no real-time visibility for facilities staff. SmartBin replaced guesswork with data.

A deployed IoT smart waste monitoring system built for PT Persita's stadium facility. Two-sensor architecture on an ESP8266 microcontroller: Sensor A monitors fill level in real time (0–100% from ultrasonic distance, 5-point rolling average for noise reduction), Sensor B triggers a touchless servo lid via hand-proximity detection. RGB LED status display (green/yellow/red by threshold). Real-time Telegram Bot notifications with 10-second debouncing, automatic WiFi reconnection, and 3-attempt message retry logic. Built across two full prototype iterations and piloted on a staffed campus reception desk before the venue install, so the unit was validated in continuous real-world use rather than bench conditions. Final system ran 72 hours continuously.

My Role

IoT Systems Developer

Responsibilities

Designed the complete hardware architecture (ESP8266 microcontroller, dual HC-SR04 ultrasonic sensors, servo motor, RGB LED array) and produced the full circuit and pin configuration documentation
Wrote all firmware in Arduino C++ including fill-level monitoring, touchless lid automation, Telegram Bot cloud integration, and automatic WiFi reconnection logic
Conducted two full prototype iterations with 3 real users in the live stadium environment, collecting structured feedback and implementing targeted fixes, moving from an unstable initial prototype to a final build that ran 72 hours continuously

Key decisions

Implemented a 5-point rolling average on Sensor A readings to reduce noise from angle variation and bin shape; this single change turned an unstable fill-level reading into a steady one
Added 10-second debouncing on Telegram notifications after initial testing revealed message spam from sensor fluctuations was frustrating users and eroding trust in the system

What I Built

Deliverables shipped

Fill level monitoring: ultrasonic distance → percentage conversion, 5-point rolling average, 3-state RGB LED (green ≤49%, yellow 50–79%, red ≥80%)
Touchless lid automation: proximity detection via Sensor B, servo motor 0°–180°, non-blocking millis() timing
Telegram Bot alerting: real-time push notifications on state change, 10-second debounce, 3-attempt retry, WiFi status included in message
Reliability systems: automatic WiFi reconnection every 30 seconds, graceful degradation, error logging

Visuals