Edge-cloud IoT roof control system combining ESP32-S3 local sensors (DHT22, BME280, rain sensor) with Firebase backend and BMKG weather API for intelligent rain prediction. Local sensor fusion triggers immediate safety responses (roof closure, buzzer alarm) when humidity exceeds 85% or rain is detected, while cloud-based Random Forest model processes regional weather data for proactive automation. This hybrid architecture minimizes latency-critical decisions at the edge while leveraging cloud ML for predictive accuracy, demonstrating best practices in IoT resilience and fail-safe design. Fork this diagram on Diagrams.so to customize sensor types, adjust prediction thresholds, or integrate alternative weather APIs and ML frameworks. The dual-layer decision logic—local rules for safety, cloud predictions for optimization—exemplifies production IoT patterns balancing responsiveness and intelligence.