A data-driven SmartInterior operates on a continuous feedback loop: IoT sensors collect environmental and behavioral data, edge gateways process it locally for privacy compliance, cloud analytics identify patterns, and automation rules self-optimize based on learned preferences. The core sensor suite includes occupancy detection (PIR or 24 GHz mmWave radar), air quality (CO2 ppm, PM2.5 μg/m³, TVOC ppb), energy metering (per-circuit clamp meters), and ambient conditions (temperature, humidity, lux levels). Data is collected at 1–5 minute intervals and stored in a time-series database — InfluxDB on a local NAS or Home Assistant’s built-in SQLite recorder — with a 90-day rolling window for pattern analysis.
Pattern recognition transforms raw data into actionable insights. After 30 days of occupancy logging, the system identifies peak usage hours per room (adjusting HVAC pre-cooling schedules 30 minutes before habitual arrival times), preferred temperature ranges per zone, and correlations between outdoor PM2.5 spikes and indoor air purifier activation. Bangkok’s dry season averages 25–40 μg/m³ PM2.5, with peaks exceeding 75 μg/m³ during January–March when the system shifts from reactive to proactive purifier activation. Thailand’s Netpie MQTT platform, developed by NECTEC, enables device-to-cloud data flows while keeping sensitive occupancy data on local storage under PDPA compliance.
KPI benchmarks show measurable improvements after a 90-day optimization cycle: thermal comfort satisfaction improves 15–25% (measured via weekly HA dashboard surveys), energy consumption drops 12–18% as scheduling aligns with actual occupancy rather than fixed schedules, and air quality alert response shifts from reactive to proactive — the system activates purifiers 10–15 minutes before PM2.5 thresholds are breached. HappySmart’s monthly reporting on these KPIs provides transparent ROI documentation for clients comparing smart home investment against baseline utility costs.