Introduction
At SWITCH, we understand that diverse transportation options are crucial for meeting the mobility needs of people. From cars to micromobility, each mode presents unique challenges. Recently, we turned our attention to the short-term car rental industry, where accurate car rental demand prediction is critical, especially at airports where demand fluctuates due to flight schedules, weather, and local events. Using PULSE-AI, we showcased how businesses can anticipate and manage demand, ensuring optimal vehicle availability and allocation. Our car rental management software, integrated into Urbiverse and Urban Copilot, equips companies with AI-CoPilot capabilities for advanced fleet management and planning.
Our advanced AI Simulation Engine, PULSE-AI, powers both Urbiverse and Urban Copilot, creating digital twins of any fleet and modeling variables like vehicle costs, autonomy, and fuel types. This system is designed to be universally applicable across industries and vehicle types, ensuring efficient fleet and resource management regardless of the sector.
Project Overview
Recently, we applied PULSE-AI to a forecasting experiment that predicted car rental trips departing daily from four major European airports. The goal was to optimize fleet management and resource allocation, demonstrating the universal application of our existing technology in the car rental industry without requiring custom development.
Challenges in Short-Term Car Rentals Management Software
Managing fleets with car rental management software requires accurate demand forecasting, purchasing, and resource allocation to meet dynamic customer demands. These challenges impact the profitability and operational efficiency of rental companies.
Key Considerations
- Vehicle Acquisition: Companies must plan their major vehicle purchases well in advance, typically around 70% of their fleet between September and October, with adjustments made throughout the year through smaller, more flexible spot purchases. Car rental demand prediction is essential to avoid over- or under-purchasing, which can lead to financial strain.
- Demand Forecasting and Allocation: To remain competitive, companies need to continuously assess demand and optimize their vehicle distribution. This involves analyzing historical data, economic trends, and external factors like seasonality, which often shift rapidly.
- Strategic Planning: Rental companies must develop business plans months in advance, adjusting throughout the year to account for evolving market conditions. Regional market differences, such as lower demand in Europe compared to the US, add further complexity.
- Purchasing Models: Striking the right balance between buyback agreements and outright vehicle purchases is critical. Over-reliance on one model can expose companies to unnecessary risks, either through market fluctuations or restricted flexibility.
These challenges highlight the need for precise forecasting and dynamic decision-making to optimize fleet management and ensure long-term financial success.
The Experiment
We targeted four key European airports:
- Frankfurt am Main Airport
- Adolfo Suárez Madrid–Barajas Airport
- Leonardo da Vinci–Fiumicino Airport
- Stockholm Arlanda Airport
Using data from two industry-leading car rental providers, we conducted a forecasting experiment across multiple time horizons—7, 30, 90, and 180 days—powered by our PULSE-AI engine.
Forecasting Results
Our models demonstrated strong predictive performance for car rental demand prediction:
- Short-Term Accuracy: The models excelled at short-term forecasts (7-day horizon), delivering highly reliable predictions with minimal error.
- Long-Term Stability: Even at longer horizons (180 days), the models maintained consistent accuracy, controlling for errors effectively.
- Actionable Insights: Our models provided granular insights into different car types and providers, enabling better-informed decisions about fleet composition and allocation.
Model Performance Metrics
We assessed our forecasting models using the following key metrics:
- Average Daily Relative Accuracy (ADRA): Measures daily forecast accuracy, especially for shorter timeframes.
- Average Monthly Relative Accuracy (AMRA): Evaluates how accurate the forecasts are over a monthly period.
- Total Number of Trips Relative Accuracy (TRA): Compares the predicted number of trips with the actual number, showing how close the forecast is to reality.
Performance Breakdown
- 1-Week Horizon: The models delivered excellent short-term accuracy, with an ADRA of 91.12%, meaning daily forecasts were highly reliable. The AMRA was 94.82%, indicating strong monthly accuracy, and the TRA was 98.04%, showing near-perfect accuracy in predicting the total number of trips.
- 1-Month Horizon: Over the 1-month period, the models maintained strong performance, with an AMRA of 90.39% and a TRA of 96.78%, showing consistent accuracy in both monthly forecasts and total trips.
- 3-Month Horizon: At this horizon, the model’s performance remained solid, with an AMRA of 87.72% and a TRA of 94.30%, reflecting a high degree of accuracy even in medium-term forecasts.
- 6-Month Horizon: Over the longest period, the model achieved an AMRA of 87.99% and a TRA of 93.12%, demonstrating stable accuracy even at longer-term horizons.
These results demonstrate not only high accuracy but also the ability of PULSE-AI to deliver consistent performance across varying time frames, reinforcing the system’s robustness for both short- and long-term forecasting.
Application of PULSE-AI for Car Rental Management Software
The success of this experiment reinforces PULSE-AI’s versatility. This AI engine is designed to be applied across industries, ensuring seamless adaptability for any fleet management challenge. Here’s how it can be applied in the Car Rental industry:
- Fleet Allocation:
- Using real-time data, our car rental management software dynamically adjusts vehicle placement to meet predicted demand, minimizing idle time and maximizing resource efficiency.
- Demand Forecasting:
- Companies can plan future vehicle purchases and optimize rental or sale decisions based on accurate demand prediction
By integrating these capabilities, PULSE-AI enables car rental companies—and other mobility providers—to improve profitability and customer satisfaction while maintaining sustainability and operational efficiency.
Urbiverse and Urban Copilot as Car Rental Fleet Management Software
PULSE-AI powers our Urban Copilot and Urbiverse platforms, that provide comprehensive fleet management and planning solutions:
- Urban Copilot offers:
- Real-Time Demand Forecasting
- Intelligent Fleet Rebalancing
- Advanced Routing Algorithms for Optimal Fleet Performance
- Mobile Application Integration for Real-Time Updates
- Urbiverse offers:
- Mobility Simulation for future infrastructure and fleet scenarios
- Autonomous Vehicle Demand Forecasting
- Fleet Eletrification Simulation
- Optimized Mobility / Logistic Hubs and Charging Station Placement
These tools are how our clients interact with our AI-Engine. If PULSE-AI is OpenAI o1, Urbiverse and Urban CoPilot are our ChatGPT.
Conclusion
his project demonstrates the remarkable potential and adaptability of PULSE-AI. By using our existing technology for car rental demand prediction, we’ve shown how PULSE-AI can seamlessly be applied to different industries without requiring custom solutions. Its ability to consistently deliver accurate forecasts across various scenarios highlights its value as a versatile, reliable tool.
With PULSE-AI, Urban Copilot, and Urbiverse, companies can:
- Transform their car rental operations through data-driven decision-making.
- Improve demand prediction accuracy, optimizing resource allocation.
- Increase profitability by enhancing fleet management strategies.
We invite car rental companies and mobility providers to partner with us and experience how our car rental management software can drive growth and efficiency across a range of industries.
Ready to revolutionize your fleet management? Contact us today to discover how SWITCH can enhance your operations with our AI-powered solutions for car rental and beyond.
