Welcome to my IoT development portfolio. This repository showcases my technical progression from bare-metal microcontroller interfaces to closed-loop automation and global cloud telemetry systems, completed during the DecodeLabs Industrial Training program.
- Goal: Connect a physical/simulated microcontroller to a digital sensor to read and process ambient environmental data.
- Core Skills: GPIO configuration, 1-Wire protocol, I2C bus communication, asynchronous timing (
millis()). - Tech Stack: ESP32, DHT22 Sensor, 16x2 I2C LCD Screen, C++ (Arduino IDE).
- Key Features:
- Implemented non-blocking mathematical delta checks (
millis()) to sample data every 2000ms. - Eliminated the self-heating sensor trap and busy-waiting CPU locks caused by static
delay()loops. - Formatted floating-point telemetry strings using static memory allocation (
dtostrf) to prevent heap fragmentation.
- Implemented non-blocking mathematical delta checks (
- Goal: Build an autonomous, closed-loop feedback system that activates mechanical hardware based on real-time environmental inputs.
- Core Skills: Analog-to-Digital Conversion (ADC), Exponential Moving Average (EMA) filtering, threshold gating with hysteresis, galvanic isolation.
- Tech Stack: ESP32 / Atmel SAM D21, Analog Soil Moisture Sensor, 5V Relay Module, DC Water Pump.
- Key Features:
- Implemented an EMA software filter to smooth out high-frequency switching noise generated by the ESP32's Wi-Fi radio.
- Designed a multi-threshold Hysteresis Deadband (ON at <30%, OFF at >45%) to completely eliminate destructive relay chattering.
- Utilized an optocoupler for galvanic isolation, safely shielding sensitive silicon logic from high-voltage inductive Back EMF spikes.
- Goal: Broadcast real-time edge physical security telemetry across the internet to a centralized cloud dashboard.
- Core Skills: Wi-Fi stack configuration, MQTT/HTTP telemetry protocols, cloud dashboard integration.
- Tech Stack: ESP32, HC-SR04 Ultrasonic Distance Sensor, MQTT Broker (Adafruit IO / HiveMQ).
- Key Features:
- Configured a robust local Wi-Fi network stack with active error-handling and auto-reconnect logic.
- Achieved ultra-low bandwidth overhead (2-byte fixed headers) by routing telemetry streams through an MQTT publish-subscribe broker instead of verbose HTTP requests.
- Built real-time edge intrusion detection utilizing time-of-flight acoustic ranging.
- Microcontrollers & Hardware: ESP32, Atmel SAM D21, Relays, CMOS 3.3V/5V logic interfacing.
- Sensors & Actuators: DHT22 (Temp/Humidity), HC-SR04 (Ultrasonic), Soil Moisture, DC Pumps.
- Networking & Protocols: MQTT, HTTP, I2C, 1-Wire, Wi-Fi Stack configuration.
- Software Architecture: Non-blocking state machines, EMA data filtering, Static memory safety.
Developed during the DecodeLabs Industrial Training Kit (Batch 2026).