Washer & Dishwasher Cycle Detection System

Copyright (c) 2025 the_louie

This AppDaemon app monitors washer and dishwasher power consumption to accurately detect when wash cycles start and end, sending notifications accordingly.

How It Works

The system analyzes power consumption patterns from your washer and dishwasher to distinguish between:

• Idle state: 0.1-1.2 watts (standby power)
• Active phases: 10-350+ watts (washing, spinning, etc.)
• Heating phases: 2000+ watts (drying, sanitizing)
• Pause periods: Brief drops to idle levels between phases
• Cycle completion: Sustained return to idle levels

Key Features

1. Trend Analysis: Uses multiple power readings to detect sustained changes, not just spikes
2. False Positive Prevention: Waits for stable idle periods before declaring cycle end
3. Double Stop Protection: Handles false stops in dishwasher cycles (unique to dishwashers)
4. Real-time Monitoring: Responds to power changes immediately via state listeners
5. Fallback Polling: Periodic checks ensure no events are missed
6. Configurable Thresholds: Adjustable power levels and timing parameters
7. Cycle Prediction: Predicts cycle end times based on historical data analysis
8. Smart Notifications: Sends notifications for cycle start, end, and predictions

Installation

1. Copy i1_washer_watcher.py and i1_disher_watcher.py to your AppDaemon apps directory
2. Copy the configuration from config.yaml to your AppDaemon configuration
3. Restart AppDaemon

Configuration

Basic Configuration

# Washer Configuration
washer_watcher:
  module: i1_washer_watcher
  class: WasherWatcher
  washer_entity: sensor.washer_power
  persons:
    - name: louie
      notify: mobile_app_iphone_28

# Dishwasher Configuration
disher_watcher:
  module: i1_disher_watcher
  class: DisherWatcher
  disher_entity: sensor.disher_power
  persons:
    - name: louie
      notify: mobile_app_iphone_28

Advanced Configuration

washer_watcher:
  module: i1_washer_watcher
  class: WasherWatcher

  # Power sensor entity
  washer_entity: sensor.washer_power

  # Detection thresholds (watts)
  idle_threshold: 1.0          # Power level considered idle
  active_threshold: 10.0       # Power level considered active
  cycle_start_threshold: 15.0  # Power level to trigger start detection

  # Timing parameters
  check_interval: 30           # Check frequency (seconds)
  trend_window: 5              # Readings for trend analysis
  stable_idle_time: 300        # Idle time before cycle end (seconds)

  # Notifications
  persons:
    - name: louie
      notify: mobile_app_iphone_28
    - name: spouse
      notify: notify.telegram

disher_watcher:
  module: i1_disher_watcher
  class: DisherWatcher

  # Power sensor entity
  disher_entity: sensor.disher_power

  # Detection thresholds (watts) - LOOSER SENSITIVITY
  idle_threshold: 1.2          # Power level considered idle (lower to catch more cycles)
  active_threshold: 12.0       # Power level considered active (lower to catch more cycles)
  cycle_start_threshold: 15.0  # Power level to trigger start detection (lower to catch more cycles)

  # Timing parameters
  check_interval: 30           # Check frequency (seconds)
  trend_window: 10             # Readings for trend analysis
  stable_idle_time: 600        # Idle time before cycle end (10 minutes - shorter for disher)
  min_cycle_duration: 480      # Minimum cycle duration (8 minutes - shorter for disher)
  cooldown_time: 600           # Cooldown between cycles (10 minutes - shorter for disher)

  # Notifications
  persons:
    - name: louie
      notify: mobile_app_iphone_28
    - name: spouse
      notify: notify.telegram

Detection Logic

Washer Cycle Detection

Cycle Start Detection

A washer cycle is considered started when:

1. Power exceeds cycle_start_threshold (default: 15W)
2. Power is significantly higher than idle (3x idle_threshold)
3. Power remains above active_threshold for 3 consecutive readings

Cycle End Detection

A washer cycle is considered ended when:

1. Power drops to idle_threshold or below
2. Stays idle for stable_idle_time (default: 5 minutes)
3. No active periods detected during the idle time
4. Had a previous active cycle (prevents false starts)

Dishwasher Cycle Detection

Cycle Start Detection

A dishwasher cycle is considered started when:

1. Power exceeds cycle_start_threshold (default: 15W) - LOOSER THRESHOLD
2. Power spike detection (2.5x recent average)
3. Sustained high readings (3+ readings above 12W)
4. Validation score of 6+ points (reduced from 8)

Cycle End Detection

A dishwasher cycle is considered ended when:

1. Power drops to idle_threshold or below
2. Stays idle for stable_idle_time (default: 10 minutes) - SHORTER TIME
3. Double stop protection: Ignores false stops within 10 minutes
4. Minimum cycle duration met (8 minutes minimum)
5. No recent high power readings

Double Stop Protection

Dishwashers have a unique issue where they can appear to stop mid-cycle:

• False stops: Brief drops to idle during heating phases
• Protection: System waits 10 minutes before considering a real stop
• Multiple false stops: Can handle up to 3 false stops per cycle
• Resume detection: Automatically resumes cycle tracking if activity continues

Historical Data Analysis

Based on analysis of your appliance logs, the system handles:

Washer Characteristics

• Power ranges: 0.1W (idle) to 2000+W (heating/spinning)
• Pause periods: Brief drops to idle between phases
• Variable cycle lengths: 30 minutes to several hours
• Multiple phases: Washing, rinsing, spinning, etc.

Dishwasher Characteristics

• Power ranges: 0.1W (idle) to 2000+W (heating/drying)
• Lower overall consumption: 10-60W during washing phases
• Heating phases: 2000+W during drying/sanitizing
• False stops: Brief pauses between heating phases
• Longer cycles: 2-6 hours typical duration

Cycle Types Identified

Washer Cycle Types

Long Heating Cycle (544.5 ± 137.3 minutes):

• Duration: 179-719 minutes (3-12 hours)
• Characteristics: Includes heating phase (2000+ watts) and high spin phase (300+ watts)
• Typical use: Heavy loads, hot water cycles, sanitize cycles
• Prediction accuracy: ±137 minutes (2+ hours)

Dishwasher Cycle Types

Standard Wash Cycle (120 minutes typical):

• Duration: 60-180 minutes (1-3 hours)
• Characteristics: Includes heating phase (2000+ watts) and high spin phase (100+ watts)
• Typical use: Normal dishwashing with drying
• Prediction accuracy: Based on historical data

Quick Wash Cycle (60 minutes typical):

• Duration: 30-90 minutes (0.5-1.5 hours)
• Characteristics: Lower power consumption, no heating phase
• Typical use: Light loads, quick cleaning
• Prediction accuracy: Based on historical data

Troubleshooting

False Start Notifications

If you get false start notifications:

• Washer: Increase cycle_start_threshold and active_threshold
• Dishwasher: Increase cycle_start_threshold (but keep lower than washer)
• Check for other appliances on the same circuit

False End Notifications

If you get false end notifications:

• Washer: Increase stable_idle_time, decrease idle_threshold
• Dishwasher: Increase stable_idle_time, check for false stops
• Check if appliance has long pause periods

No Notifications

If you don't get notifications:

• Verify washer_entity/disher_entity points to correct sensor
• Check notification service configuration
• Review AppDaemon logs for errors

Dishwasher-Specific Issues

Missing cycles:

• The dishwasher uses looser thresholds to catch more cycles
• If still missing cycles, decrease thresholds further
• Check for very brief cycles that might be below minimum duration

False stops detected as real ends:

• The double stop protection should handle this automatically
• If issues persist, increase false_stop_cooldown time

Logging

The app logs all detection events and errors. Check your AppDaemon logs for:

• Cycle start/end detection
• Power readings and trends
• False stop detection (dishwasher only)
• Cycle type identification and predictions
• Notification delivery status
• Configuration errors

Prediction Features

Cycle End Prediction

The system predicts cycle end times based on historical data analysis:

1. Automatic Detection: Identifies cycle type when heating phase is detected
2. Real-time Updates: Continuously monitors power patterns during cycle
3. Smart Notifications: Sends alerts when:
   • Cycle is finishing soon (within 5 minutes)
   • Cycle is overdue (taking longer than predicted)
4. Accuracy: Predictions improve with more historical data

Notification Types

Washer Notifications

• 🧺 Cycle Start: When a new washer cycle begins
• ✅ Cycle End: When washer cycle completes successfully
• ⏰ Finishing Soon: When cycle is predicted to end within 5 minutes
• ⚠️ Overdue: When cycle is taking longer than expected

Dishwasher Notifications

• 🍽️ Cycle Start: When a new dishwasher cycle begins
• ✅ Cycle End: When dishwasher cycle completes successfully
• ⏰ Finishing Soon: When cycle is predicted to end within 5 minutes
• ⚠️ Overdue: When cycle is taking longer than expected

Data Structure

The system stores historical cycle data in JSON format for machine learning predictions:

Washer Data: washer_historical_cycles.json Dishwasher Data: disher_historical_cycles.json

Each cycle record contains:

• Start/end times
• Duration in minutes
• Maximum power consumption
• Detected phases (heating, high spin)
• Cycle type classification
• Prediction accuracy metrics

The system processes power readings in this format:

YYYY-MM-DD HH:MM:SS [device] state: [state] power: [old_value] -> [new_value]

Example:

2025-07-06 13:11:23 washer state: running power: 27.2 -> 21.4
2025-07-06 14:30:15 disher state: running power: 15.8 -> 12.3

Files

• i1_washer_watcher.py - Main washer AppDaemon app
• i1_disher_watcher.py - Main dishwasher AppDaemon app
• config.yaml - Configuration example for both appliances
• washer_historical_cycles.json - Historical washer cycle data (ML)
• disher_historical_cycles.json - Historical dishwasher cycle data (ML)

License

This project is licensed under the BSD 2-Clause License - see the LICENSE file for details.

Copyright (c) 2025 the_louie