Augmented Smart Home Dataset with Weather Information

Dataset Overview

This augmented smart home dataset significantly expands upon the original work by Taranvee, providing comprehensive household energy monitoring at one-minute resolution with enhanced device coverage and environmental context. Published on Mendeley Data in February 2025, it represents realistic residential IoT deployments.

Enhanced Device Coverage - 13+ Appliances

The dataset tracks energy consumption in kilowatts (kW) for a diverse range of household devices:

Original Appliances

• Kitchen: Dishwasher, Refrigerator, Microwave
• Work Spaces: Home Office equipment
• Climate Control: HVAC systems

Augmented IoT Devices (New)

• Electric Vehicle: Car charger with high power consumption patterns
• Water Systems: Water heater with thermal cycling behavior
• Climate Systems: Dedicated air conditioning monitoring
• Entertainment: Home theater system power profiles
• Outdoor Infrastructure: Smart outdoor lighting
• Utility: Laundry machines (washer/dryer)
• Recreation: Pool pump with scheduled operation patterns

Weather Integration

Regional environmental data provides crucial context for energy consumption analysis:

• Temperature (°C): Outdoor temperature affecting heating/cooling loads
• Humidity (%): Relative humidity impacting HVAC efficiency
• Atmospheric Pressure (hPa): Barometric conditions
• Weather Conditions: Categorical data (sunny, cloudy, rainy) affecting lighting and appliance usage

Temporal Resolution

The dataset offers one-minute granularity, enabling detailed analysis of:

• Appliance startup transients and power spikes
• Daily usage patterns and peak demand periods
• Correlation between weather events and energy consumption
• Load forecasting for demand response programs
• Disaggregation of mixed appliance signals (NILM research)

Research Applications

Energy Management

• Load forecasting and demand prediction models
• Peak shaving and demand response optimization
• Renewable energy integration planning
• Time-of-use tariff analysis

Machine Learning Tasks

• NILM: Non-Intrusive Load Monitoring (appliance disaggregation)
• Anomaly Detection: Identify faulty appliances or unusual consumption
• Occupancy Detection: Infer home occupancy from appliance patterns
• Behavior Modeling: Understand resident activity patterns
• Weather Correlation: Regression models linking climate and energy use

Smart Grid Research

• Distributed energy resource management
• Virtual power plant aggregation
• Electric vehicle charging optimization
• Grid stability and voltage regulation studies

Data Quality

The dataset benefits from smart meter accuracy, continuous monitoring without gaps, synchronized timestamps across all devices, and validated weather data from regional meteorological stations. The one-minute resolution strikes optimal balance between detail and data volume.

Multi-Device Scenarios

The expanded device list enables research into complex, multi-device smart home scenarios including simultaneous EV charging and HVAC operation, coordinated lighting and entertainment system usage, and interaction between pool pumps and solar generation.