Sound-Based Detection for Maytag Washing Machine | Generated by AI

Home PDF

Using sound to detect when your Maytag MHW8630HW washing machine stops is a viable alternative to the camera-based approach. By placing a sound sensor close to the washing machine, you can monitor the noise generated during operation (e.g., motor hum, water flow, or drum rotation) and detect when it stops, indicating the end of the wash cycle. Below, I’ll explain how to adapt the previous project to use a sound sensor instead of a camera, focusing on the hardware, algorithm, and code changes while keeping the Telegram notification system.

Tech Stack

Hardware

  1. Arduino Board:
    • ESP32-CAM (still usable) – Retains Wi-Fi for Telegram integration, though the camera is not needed.
    • Alternative: ESP8266 NodeMCU or Arduino Uno with an ESP8266 module for Wi-Fi (simpler if you don’t need the camera).
  2. Sound Sensor:
    • KY-038 Microphone Sound Sensor or similar – Affordable, detects sound levels via an analog output.
    • Alternative: MAX9814 Electret Microphone Amplifier – More sensitive, with adjustable gain for better detection.
  3. Power Supply:
    • 5V USB power adapter or battery pack for the ESP32 or other board.
  4. Mounting:
    • Place the sound sensor close to the washing machine (e.g., taped to the side or top) where it can detect operational sounds but avoid direct water exposure.
    • Use a small enclosure to protect the sensor and board.
  5. Wi-Fi Router:
    • For internet connectivity to send Telegram notifications.

Software

Algorithm for Detecting Washing Machine Status with Sound

The sound sensor will detect the noise level produced by the washing machine. When the machine is running, it generates consistent sounds (e.g., motor, water, or drum). When it stops, the sound level drops significantly. The algorithm processes these sound levels to determine the machine’s status.

Detection Algorithm

  1. Sound Sampling:
    • Continuously read the analog output from the sound sensor to measure noise levels.
  2. Signal Processing:
    • Averaging: Calculate the average sound level over a short window (e.g., 1-2 seconds) to smooth out transient noises (e.g., a door slam).
    • Thresholding: Compare the average sound level to a predefined threshold. A high level indicates the machine is running, while a low level suggests it’s stopped.
  3. State Machine:
    • Track the machine’s state (ON or OFF) based on sound levels.
    • If the sound level exceeds the threshold for several cycles, assume the machine is running.
    • If the sound level drops below the threshold and remains low for a set period (e.g., 5 minutes), assume the wash cycle is complete.
  4. Debouncing:
    • Implement a delay (e.g., 5 minutes) to confirm the machine has stopped, avoiding false notifications during quiet phases (e.g., soaking or pauses in the cycle).
  5. Notification:
    • When the machine is confirmed stopped, send a Telegram message (e.g., “Washing machine stopped! Time to hang up clothes.”).

Why Sound Detection?

Implementation Guide

Step 1: Set Up the Telegram Bot

Step 2: Hardware Setup

  1. Choose a Sound Sensor:
    • KY-038: Provides an analog output (0-1023 for ESP32’s 10-bit ADC) proportional to sound intensity. It also has a digital output, but analog is better for nuanced detection.
    • MAX9814: More sensitive, with adjustable gain via a potentiometer. Connect to an analog pin.
  2. Connect the Sound Sensor:
    • For KY-038:
      • VCC to 5V (or 3.3V, depending on the board).
      • GND to GND.
      • Analog Out (A0) to an analog pin on the ESP32 (e.g., GPIO 4).
    • For MAX9814:
      • Similar connections, but adjust gain using the onboard potentiometer for optimal sensitivity.
  3. Position the Sensor:
    • Place the sensor close to the washing machine (e.g., on the side or top) where it can detect motor or drum noise. Avoid areas with water exposure.
    • Test placement by monitoring sound levels during a wash cycle (use Serial Monitor to log values).
  4. Power the Board:
    • Connect a 5V USB power adapter or battery pack to the ESP32 or other board.
    • Ensure stable power, as Wi-Fi communication requires consistent voltage.
  5. Mounting:
    • Use a small enclosure or tape to secure the sensor and board, ensuring the microphone is exposed to capture sound.

Step 3: Software Setup

Step 4: Write the Arduino Code

Below is a sample Arduino sketch for the ESP32 (or ESP8266) to detect sound levels and send Telegram notifications. This assumes a KY-038 sound sensor connected to GPIO 4.

#include <WiFi.h>
#include <UniversalTelegramBot.h>
#include <ArduinoJson.h>

// Wi-Fi credentials
#define WIFI_SSID "your_wifi_ssid"
#define WIFI_PASSWORD "your_wifi_password"

// Telegram Bot credentials
#define BOT_TOKEN "your_bot_token"
#define CHAT_ID "your_chat_id"

// Sound sensor pin
#define SOUND_PIN 4 // GPIO 4 for analog input

// Sound detection parameters
#define SOUND_THRESHOLD 300 // Adjust based on testing (0-1023)
#define SAMPLE_WINDOW 2000 // 2 seconds for averaging
#define STOP_DELAY 300000 // 5 minutes in milliseconds

WiFiClientSecure client;
UniversalTelegramBot bot(BOT_TOKEN, client);

bool machineRunning = false;
unsigned long lastOnTime = 0;

void setup() {
  Serial.begin(115200);

  // Connect to Wi-Fi
  WiFi.begin(WIFI_SSID, WIFI_PASSWORD);
  while (WiFi.status() != WL_CONNECTED) {
    delay(500);
    Serial.print(".");
  }
  Serial.println("WiFi connected");

  // Configure Telegram client
  client.setInsecure(); // For simplicity; consider proper SSL in production

  // Set up sound sensor pin
  pinMode(SOUND_PIN, INPUT);
}

void loop() {
  // Sample sound level over a window
  unsigned long startMillis = millis();
  uint32_t totalSound = 0;
  uint16_t sampleCount = 0;

  while (millis() - startMillis < SAMPLE_WINDOW) {
    totalSound += analogRead(SOUND_PIN);
    sampleCount++;
    delay(10); // Small delay between samples
  }

  float avgSound = sampleCount > 0 ? (float)totalSound / sampleCount : 0;
  Serial.print("Average sound level: ");
  Serial.println(avgSound);

  // State machine
  if (avgSound > SOUND_THRESHOLD) {
    if (!machineRunning) {
      machineRunning = true;
      Serial.println("Machine is ON");
    }
    lastOnTime = millis();
  } else {
    if (machineRunning && (millis() - lastOnTime > STOP_DELAY)) {
      machineRunning = false;
      Serial.println("Machine stopped");
      bot.sendMessage(CHAT_ID, "Washing machine stopped! Time to hang up clothes.", "");
    }
  }

  delay(10000); // Check every 10 seconds
}

Step 5: Customize the Code

  1. Update Credentials:
    • Replace your_wifi_ssid, your_wifi_password, your_bot_token, and your_chat_id with your actual values.
  2. Tune SOUND_THRESHOLD:
    • Run the washing machine and monitor sound levels via the Serial Monitor (Serial.println(analogRead(SOUND_PIN));).
    • Set SOUND_THRESHOLD to a value above ambient noise but below the machine’s operational noise (e.g., 200-500, depending on your setup).
  3. Adjust SAMPLE_WINDOW:
    • A 2-second window (2000 ms) smooths out transient noises. Increase if background noise causes false readings.
  4. Adjust STOP_DELAY:
    • Set to 300000 (5 minutes) to avoid false notifications during quiet phases like soaking.

Step 6: Test and Deploy

  1. Upload the Code:
    • Connect the ESP32 to your computer via a USB-to-serial adapter.
    • Select ESP32 Wrover Module (or NodeMCU for ESP8266) in Arduino IDE and upload the sketch.
  2. Test the System:
    • Start the washing machine and monitor the Serial Monitor for sound levels and state changes.
    • Verify Telegram notifications when the machine stops.
  3. Fine-Tune:
    • Adjust SOUND_THRESHOLD or STOP_DELAY if false positives/negatives occur.
    • Test in different conditions (e.g., with background noise) to ensure reliability.
  4. Permanent Installation:
    • Secure the sensor and board in an enclosure near the machine, ensuring the microphone is exposed but protected from water.

Advantages of Sound Detection

Challenges and Mitigations

Optional Enhancements

Troubleshooting

This sound-based approach is simpler and more robust in varying light conditions compared to the camera-based method, making it ideal for a washing machine monitor. Let me know if you need help with specific steps, sensor calibration, or integrating additional features!

Back 2025.06.28 Donate