Visualizing Smart Water Systems with AI & GIS

Apr 30, 2025·
· 1 min read
Image credit: Unsplash

In my recent work on hybrid AI models for water infrastructure, I needed ways to effectively explain system behavior, learning outcomes, and detected anomalies. Here are tools I used with HugoBlox to communicate those insights interactively and clearly.

Interactive Charts with Plotly

I used Plotly for dynamic visualizations of:

  • Predicted vs actual pressure across nodes
  • Leak probability heatmaps over time
  • Consumption anomalies from SCADA time series

Save a Plotly .json file (e.g., leak-probability.json) in your page folder and display it like this:











graph TD
A[SCADA Data] --> B(GNN Inference)
B --> C{Leak Detected?}
C -->|Yes| D[Trigger Alert]
C -->|No| E[Continue Monitoring]


sequenceDiagram
SCADA->>GNN: Feed pressure data
GNN->>PINN: Pass edge/node outputs
PINN-->>SCADA: Simulated parameters
Note right of PINN: Feedback to SCADA control
AI for Water GNN SCADA Plotly GIS HugoBlox
Authors
GeoAI & Machine Learning Engineer
Hybrid AI researcher working with GNN, PINN, SCADA, and water infrastructure.