How sukrut10k/smart-water-leak-detection works
This is a Streamlit water-leak monitoring dashboard backed by machine-learning models. Users can enter sensor readings or upload a CSV, then inspect predictions, risk simulations, geographic views, and analytics; an AI assistant provides dataset-aware insights. The notebook is documented as the model-training and evaluation workflow. The excerpts show much of the dashboard behavior, but omit some surrounding code, including dataset-loading details, so uncertain data wiring is not shown.
Main entry points: app.py Streamlit dashboard · water_leak_detection.ipynb Trained models
Components 12
Dashboard workflows
- Streamlit dashboard
frontend/app.py - Leak prediction
frontend/app.py - What-if analysis
frontend/app.py - Batch prediction
frontend/app.py - Zone map
frontend/app.py - Analytics insights
frontend/app.py - AI assistant
frontend/app.py
Model workflow
- Trained models
ml_model/water_leak_detection.ipynb
Datasets
- GIS leakage data
data/location_aware_gis_leakage_dataset.csv - Testing data
data/testing.csv
External services
- Groq API
Other components
- Dashboard user
Connections 11
- Dashboard user → Streamlit dashboard: uses
- Streamlit dashboard → Leak prediction: dispatches
- Streamlit dashboard → What-if analysis: dispatches
- Streamlit dashboard → Batch prediction: dispatches
- Streamlit dashboard → Zone map: dispatches
- Streamlit dashboard → Analytics insights: dispatches
- Streamlit dashboard → AI assistant: dispatches
- Leak prediction → Trained models: predicts with
- What-if analysis → Trained models: predicts with
- Batch prediction → Trained models: predicts with
- AI assistant → Groq API: requests insights