A data-driven system to dynamically adjust parking prices based on real-time demand factors like occupancy, queue length, traffic, and competition. The goal is to improve parking utilization and reduce congestion in urban areas.
- 📈 Dynamic pricing based on demand
- ⚡ Real-time-like streaming simulation
- 🧠 Multi-model pricing strategy
- 🌍 Competitive pricing using location
- 📊 Data visualization and analysis
- Price increases with occupancy
- Simple and stable baseline
Uses:
- Occupancy rate
- Queue length
- Traffic
- Special day
- Vehicle type
📌 Pricing formula:
Price = BasePrice × (1 + λ × Demand)
- Uses latitude & longitude
- Adjusts price based on nearby parking lots
- Prevents overpricing and improves competitiveness
- Model 1 avg price: ~12.55
- Model 2 avg price: ~11.99
- Model 2 shows higher responsiveness
- Avg occupancy: ~50%
- Peak occupancy: >90%
- Queue avg: ~4.6 vehicles
- Model 1: 10.5 → 14.7
- Model 2: 10.0 → 15.0
- Smooth variations, no extreme spikes
- Higher demand → higher price
- Queue = unmet demand signal
- Traffic reduces accessibility
- Competitive pricing improves balance
Capstone-Project/
│
├── Dynamic_Pricing_Parking.ipynb
├── dataset.csv
├── results/
│ ├── model1_prices.csv
│ └── model2_prices.csv
├── README.md
└── problem statement.pdf
- Python
- Pandas, NumPy
- Pathway
- Bokeh / Matplotlib
- Open notebook in Google Colab
- Upload dataset.csv
- Run all cells
- View pricing outputs and plots
- Real-time deployment
- ML-based demand prediction
- Reinforcement learning pricing
- Smart rerouting system
Sarthak Singh IIEST Shibpur
