Machine Maintenance Monitoring System

An IIoT-based machine condition monitoring node using ESP32, AHT30 (temperature & humidity), and MPU6050 (3-axis vibration). Designed for predictive maintenance in smart factory environments with ultra-low power consumption for long-term battery operation.

The Fast Fourier Transform (FFT) serves as the mathematical core of the system, transforming complex time-domain vibration data into a clear frequency spectrum to act as a diagnostic "microscope" for pinpointing specific mechanical failures.

The node eliminates on-device displays entirely — all monitoring is handled by a C# DevExpress dashboard over WiFi TCP, following industrial IoT conventions.

---

Final Product

overview.mp4

Demo Overview

demo.mp4

---

Features

• Real-time vibration monitoring — 3-axis acceleration (X/Y/Z) via MPU6050
• Environmental monitoring — Temperature & humidity via AHT30
• Battery monitoring — Voltage and percentage via ADC voltage divider (GPIO 34)
• Deep Sleep + Motion Interrupt — ESP32 wakes only on vibration (EXT0) or 5-minute timer
• LED status indicators — Red / Yellow / Green LEDs for WiFi connection feedback
• Magnetic mounting — Neodymium magnets embedded in 3D-printed enclosure for tool-free attachment to metal surfaces
• WiFi TCP transmission — JSON data streamed to C# DevExpress dashboard
• Anomaly detection — Configurable WARN / ALARM thresholds on vibration magnitude
• Wake statistics — Tracks total wakes vs. motion-triggered wakes for anomaly frequency analysis

---

System Architecture

┌─────────────────────────────────────────────────────────┐
│                  ESP32-WROOM-32U                        │
│                                                         │
│  AHT30   ──I2C──► Temp / Humidity                       │
│  MPU6050 ──I2C──► Accel X/Y/Z  ──► Motion INT (EXT0)    │
│  ADC Divider ───► GPIO 34  ──► Battery Voltage/Percent  │
│                                          │              │
│  Deep Sleep ◄────────────────────────────┘              │
│       │                                                 │
│       └──► WiFi TCP ──► JSON ──► C# Dashboard           │
└─────────────────────────────────────────────────────────┘

---

Hardware

Component	Description
ESP32-WROOM-32U	Main MCU with external antenna
AHT30	Temperature (±0.3°C) & Humidity (±2% RH), I2C 0x38
MPU6050	3-axis accelerometer + gyroscope, I2C 0x68/0x69 (auto-detected)
2× 18650 Li-ion (2S)	Series pack, ~7.4V, ~3000mAh
2-Series Charging	Type-C, 6.0V(0%) - 8.4V(100%)
MP1584EN buck converter	Steps 2S pack down to stable 3.3V
Voltage divider (22kΩ / 10kΩ)	Scales battery voltage to ADC-safe range
3× status LEDs	Red (fail) GPIO 27 / Yellow (connecting) GPIO 26 / Green (success) GPIO 14

---

Wiring

AHT30 / MPU6050          ESP32-WROOM-32U
────────────────         ───────────────
VCC             ──────►  3.3V
GND             ──────►  GND
SDA             ──────►  GPIO 21
SCL             ──────►  GPIO 22
MPU6050 INT     ──────►  GPIO 33    ← Motion wakeup (EXT0)
MPU6050 AD0     ──────►  GND        ← I2C address 0x68

Battery divider output ──► GPIO 34  ← ADC battery sensing
                                       (R1=22kΩ, R2=10kΩ, scale ×3.2)

LED Red    ──────────────► GPIO 27  ← WiFi fail
LED Yellow ──────────────► GPIO 26  ← WiFi connecting
LED Green  ──────────────► GPIO 14  ← WiFi success

---

JSON Output Format

The system transmits three distinct packet types: vib for burst motion data, fft for spectral analysis, and status for routine environment updates.

Vibration packet (motion wakeup — 20-sample burst)

{
  "type": "vib",
  "boot": 12,
  "idx": 1,
  "wake": "MOTION",
  "ax": "1.234",
  "ay": "0.521",
  "az": "9.812",
  "mag": "9.957",
  "status": "OK",
  "temp": 28.50,
  "hum": 65.30,
  "vbat": "8.24",
  "pbat": "93.3",
  "totalWakes": 13,
  "motionWakes": 3
  "peakFreq": 10.50
}

FFT packet (frequency spectrum data)

{
  "type": "fft",
  "fft_data":
  [
    "0.0",
    "1.5",
    "4.2",
    "12.8",
    "..." 
  ]
}

Status packet (timer wakeup — every 5 minutes)

{
  "type": "status",
  "boot": 12,
  "totalWakes": 12,
  "motionWakes": 2,
  "temp": 28.50,
  "hum": 65.30,
  "vbat": "8.24",
  "pbat": "93.3"
}

status field values: OK · WARN (≥ 2.0 m/s²) · ALARM (≥ 5.0 m/s²)

wake field values: MOTION · TIMER · RESET

---

Battery Monitoring

The 2S Li-ion pack (6.0V – 8.4V) is measured via a resistor divider that scales the voltage to the ESP32 ADC input range (0 – 3.3V).

Battery+ ──► R1 (22kΩ) ──┬──► R2 (10kΩ) ──► GND
                          └──► GPIO 34 (ADC)

Pack voltage	ADC pin voltage	Reported percentage
8.4V (full)	~2.625V	100%
7.2V (nominal)	~2.25V	50%
6.0V (empty)	~1.875V	0%

The firmware averages 10 ADC samples per reading to reduce noise.

---

Power Consumption & Battery Life

Mode	Current
ESP32 deep sleep	~10 µA
MPU6050 motion-detect mode	~20 µA
AHT30 idle	~25 µA
MP1584EN + charging circuit (sleep overhead)	~1–2 mA
Active (wake + WiFi TX)	~120 mA

Battery life calculation — 2S 18650 pack, ~3000mAh at 7.4V nominal:

Parameter	Value
Gross capacity	3000 mAh
MP1584EN efficiency	~96%
Usable capacity (at 3.3V rail)	2550 mAh
Active current (WiFi TX)	~120 mA
Active duration per wake	~3 s
Energy per wake	120 mA × (3/3600) h = 0.1 mAh
Max wake cycles	2550 / 0.1 = ~25,500 wakes

At TIMER_SLEEP_SEC = 300 (1 wake per 5 min = 288 wakes/day):

~88 days (~3 months) before recharge

Note: The MP1584EN buck converter and charging circuit draw ~1–2 mA even during deep sleep, reducing real-world runtime by roughly 10–15% compared to a theoretical bare-chip calculation.

---

Getting Started

1. Clone the repository

git clone https://github.com/phandienxauxa/Machine-Maintenance-Monitoring-System-.git
cd Machine-Maintenance-Monitoring-System-

2. Configure credentials

cp include/config.example.h include/config.h

Edit include/config.h and fill in your WiFi SSID, password, host IP, and TCP port.

3. Open in VS Code with PlatformIO

code .

PlatformIO will automatically install required libraries on first build.

4. Build & Upload

Click Build (✓) then Upload (→) in the PlatformIO toolbar, or:

pio run --target upload

5. Monitor output

pio device monitor

---

Libraries

Library	Version	Purpose
Adafruit AHTX0	^2.0.5	AHT30 sensor driver
Adafruit Unified Sensor	^1.1.14	Sensor abstraction layer
ArduinoJson	^7.0.0	JSON serialization
ArduinoFFT	^3.x.x	Fast-Fourier-Transform Processing

MPU6050 communicates directly via Wire.h with raw register access — no additional library required.

---

Configuration

Key parameters in src/main.cpp:

Parameter	Default	Description
TIMER_SLEEP_SEC	300	Timer wakeup interval (seconds)
VIB_BURST_SAMPLES	20	Samples per motion wakeup
VIB_BURST_DELAY	100	ms between burst samples
MOTION_THRESHOLD	15	MPU6050 motion sensitivity (1 unit ≈ 2mg)
MOTION_DURATION	2	MPU6050 motion duration register
THR_WARN	2.0	Warning threshold (m/s²)
THR_ALARM	5.0	Alarm threshold (m/s²)
BATTERY_MAX_VOLTAGE	8.4	2S full voltage (V)
BATTERY_MIN_VOLTAGE	6.0	2S cutoff voltage (V)

---

Project Structure

aht30_mpu6050/
├── src/
│   └── main.cpp              # Firmware — deep sleep, sensors, JSON TX
├── include/
│   ├── config.example.h      # WiFi & TCP config template (safe to push)
│   └── config.h              # Your actual credentials (gitignored)
├── platformio.ini            # PlatformIO build configuration
└── README.md

---

License

MIT License — feel free to use and modify for your own projects.