Geological hazard warning platform with GIS map and ML microservice
The project provides a complete local stack to monitor, detect, and alert on landslides, debris flows, rockfalls, and ground collapses. It combines a Cesium‑based 3D map, a point registry, MQTT device monitoring, rule‑based and machine‑learning alert pipelines, and a FastAPI microservice for model management. Designed for municipal safety teams or researchers needing an end‑to‑end demo, it offers a ready‑to‑run solution without external cloud services. Its tight integration of GIS, IoT, and ML makes it more cohesive than piecemeal tutorials.
View on GitHub →oneJue/geo-hazard-warning