This project implements a dynamic pricing model for parking spots using real-time and historical parking data.
It adjusts parking prices automatically based on multiple factors like occupancy, queue length, nearby traffic, vehicle type, special days, and competitor prices — ensuring optimal utilization and revenue.
Parking spaces in cities face dynamic demand throughout the day. Static pricing leads to inefficiencies — underutilized spots or overcrowded cheap spots.
This pipeline reads parking data, calculates baseline prices, adjusts them based on demand drivers and nearby competitors, and outputs optimal prices in real-time.
Key features:
- Parses real parking data with robust error handling.
- Uses Haversine distance to find nearby competitors.
- Dynamically adjusts prices with a competitive strategy.
- Modular, fully logged, and production-ready.
| Tech | Purpose |
|---|---|
| Python 3.x | Core scripting |
| Pathway | Data pipelines, UDFs, joins |
| pandas | (Optional for local CSV checks) |
| Logging | Debugging and audit trail |
| Mermaid | Architecture diagram |
| CSV | Input & output format |
flowchart TD
A[📂 dataset.csv] --> B[📥 Read CSV with Schema]
B --> C[🕓 Parse DateTime]
C --> D[⚙️ Baseline Price Calculation]
D --> E[📈 Demand Calculation]
E --> F[💰 Demand-based Price]
F --> G[🔀 Competitor Cross-Join]
G --> H[📏 Haversine Distance]
H --> I[🚦 Filter Nearby Competitors]
I --> J[📊 Group & Aggregate Competitor Prices]
J --> K[⚖️ Final Price Adjustment]
K --> L[📤 Output: CSV Export]
L --> M[✅ Pipeline Run]