Dynamic pricing engine for urban parking lots using real-time data streaming, demand drivers & competitive pricing — built with Python (Pandas) and Pathway applications. Includes real-time Data Visualization using Bokeh plots and Panel.
| Layer | Tools / Libraries |
|---|---|
| Streaming data processing | Pathway |
| Data manipulation | pandas, numpy |
| Visualization | Bokeh, Panel |
| Notebook & orchestration | Google Colab / Jupyter |
| Storage (CSV simulation) | pandas, local CSV files |
Goal:
- Ingest real-time parking data stream
- Compute demand dynamically based on occupancy, queue, traffic, special days & vehicle types
- Compute real competitor price using latitude & longitude
- Adjust price competitively
- Visualize daily price trend live
flowchart TD
A[Parking CSV with lat/lon] --> B[Precompute nearby lots]
B --> C[Pathway streaming]
C --> D[Tumbling window per lot]
D --> E[Self-join competitor avg price]
E --> F[Competitive price]
F --> G[Live Bokeh+Panel plot]