How sandeshnepal106/al-based-rockfall-prediction-and-alert-system-for-open-pit-mines works
The product is a research-prototype rockfall-risk dashboard. Its main workflow is image upload, location selection, latest-upload lookup, Python prediction, and map/risk display. The graph also preserves the documented model-training workflow. The README documents prediction orchestration and integrations; the sampled source does not expose the full prediction controller implementation, so that link is documented rather than inferred from unsampled code.
Main entry points: App.jsx Dashboard state · locationRoutes.js Location routes · server.js Express server · imageRoutes.js Image upload route
Components 26
Web dashboard
- Dashboard state
vite-project/src/App.jsx - Header
vite-project/src/components/header.jsx - Image upload UI
vite-project/src/components/fileupload.jsx - Image preview
vite-project/src/components/imagepreview.jsx - Location search UI
vite-project/src/components/locationsearch.jsx - Risk map
vite-project/src/components/riskmap.jsx - Map view
vite-project/src/components/map.jsx
Express API
- Express server
server/server.js - Image upload route
server/routes/imageRoutes.js - Location routes
server/routes/locationRoutes.js - Upload handler
server/controller/imageController.js - Location resolver
server/controller/locationController.js - Latest image lookup
server/middleware/getImage.js - Prediction orchestration
server/controller/getLatestInfo.js
Prediction and training
- Risk predictor
python_service/predict_rockfall.py - Trained model
ml_training/model.pkl - Model training
ml_training/train_model.ipynb - Rockfall dataset
ml_training/yosemodiFinal_with_rockfall.csv
Image storage
- Cloudinary upload
server/middleware/multer.js - Image records
server/models/imageModel.js - MongoDB connection
server/config/mongodb.js - Cloudinary client
server/config/cloudinary.js
Other components
- Site user
- Nominatim
- WeatherAPI
- Python process
Connections 29
- Site user → Dashboard state: uses
README.md - Dashboard state → Header: renders
vite-project/src/App.jsx - Dashboard state → Image upload UI: renders
vite-project/src/App.jsx - Dashboard state → Image preview: renders
vite-project/src/App.jsx - Dashboard state → Location search UI: renders
vite-project/src/App.jsx - Dashboard state → Risk map: renders
vite-project/src/App.jsx - Dashboard state → Express server: uploads and requests
vite-project/src/App.jsx - Dashboard state → Risk map: passes risk data
vite-project/src/App.jsx - Risk map → Map view: displays location
- Express server → Image upload route: mounts
server/server.js - Express server → Location routes: mounts
server/server.js - Express server → MongoDB connection: connects
server/server.js - Image upload route → Cloudinary upload: processes upload
server/routes/imageRoutes.js - Image upload route → Upload handler: dispatches to
server/routes/imageRoutes.js - Cloudinary upload → Cloudinary client: stores images
server/middleware/multer.js - Upload handler → Image records: saves record
server/controller/imageController.js - Location routes → Latest image lookup: loads latest image
server/routes/locationRoutes.js - Location routes → Location resolver: dispatches to
server/routes/locationRoutes.js - Location routes → Prediction orchestration: dispatches prediction
server/routes/locationRoutes.js - Latest image lookup → Image records: reads latest record
server/middleware/getImage.js - Location resolver → Image records: updates coordinates
server/controller/locationController.js - Location resolver → Nominatim: searches locations
server/controller/locationController.js - Location routes → Nominatim: searches suggestions
server/routes/locationRoutes.js - Prediction orchestration → Python process: runs predictor
README.md - Python process → Risk predictor: executes
README.md - Risk predictor → Trained model: loads model
README.md - Risk predictor → WeatherAPI: retrieves weather
README.md - Model training → Rockfall dataset: loads data
README.md - Model training → Trained model: serializes pipeline
README.md