RMS: Removing Barriers to Analyze the Availability and Surge Pricing of Ridesharing Services

Literature Title

RMS: Removing Barriers to Analyze the Availability and Surge Pricing of Ridesharing Services

Literature Information

• Subject Area: Sharing economy, supply and pricing dynamics analysis of ridesharing services
• Keywords: ridesharing services, sharing economy, Uber, surge pricing, COVID-19, gig economy, availability, data measurement, supply-demand analysis, dynamic pricing

Research Background and Issues

• Identified Problems or Challenges:

  • Ridesharing services (e.g., Uber and Lyft) do not disclose critical real-time data, such as availability (supply, utilization, idle time, and idle distance) and dynamic surge pricing models, limiting academic and regulatory analysis of these services.
  • Existing research data is spatially and temporally sensitive, primarily focusing on North America, specific cities, and a few service providers, resulting in poor generalizability and reproducibility.

• Significance:

  • Ridesharing services have widespread socio-economic impacts, involving transportation patterns, carbon emissions, income disparity, and traffic congestion.
  • Transparent, real-time data is crucial for policymaking and studying the global application and impact of ridesharing services.

• Research Motivation and Related Work:

  • Previous studies have explored supply and surge pricing issues using various methods, such as user surveys, driver experiences, and data scraping. However, these methods are either limited in scale or effective only for specific periods and regions, leading to a lack of universal data analysis tools.
  • Prior research has faced challenges such as overestimating supply, miscalculating utilization, and missing data due to the impact of the COVID-19 pandemic.

Solution

• Proposed Method or Solution:

  • Developed the "Ridesharing Measurement Suite" (RMS), an open-source tool capable of collecting, processing, and publicly sharing real-time data on ridesharing service availability and surge pricing.
  • RMS captures real-time data by invoking HTTPS requests from ridesharing service websites or mobile applications, processes the data, and displays it via APIs and graphical interfaces.

• Innovations:

  1. Universality: RMS supports data collection and analysis for multiple ridesharing services and is applicable to any city worldwide.
  2. Location Independence: RMS can be deployed anywhere without city-specific adjustments.
  3. Real-Time Tracking: Continuous, real-time monitoring of ridesharing services.
  4. Shareability: Provides public APIs and data-sharing mechanisms, lowering the barrier for research.

• Implementation Steps:

  • Data Collection: Acquire network traffic from ridesharing apps using Android emulators and network proxy tools; use algorithms to eliminate noise.
  • Data Processing: Extract vehicle location information, clean and analyze data (including deduplication and error elimination).
  • Data Visualization: Distribute results to users and researchers through heatmaps and JSON-formatted APIs.

• Key Technologies Used:

  • Utilized Charles proxy tool and reverse engineering methods to bypass SSL pinning and decrypt HTTPS communication.
  • Data analysis included statistical methods (e.g., Pearson correlation coefficient) and network visualization techniques (e.g., heatmaps).

Research Outcomes

• Specific Results:

  1. Developed RMS, an open tool for real-time data collection and sharing.
  2. Established a data collection network covering 10 ridesharing services across 9 countries, analyzing 8 weeks of data (from pre-pandemic and pandemic periods).
  3. Provided key insights into supply, utilization, and surge pricing impacts, such as:
     • Service supply decreased by 54% during the pandemic, while utilization increased by 6%, and surge pricing frequency rose fivefold.
     • Surge pricing was concentrated in small areas, with an impact radius of approximately 0.5 miles.
     • In multiple cities and services, over 20% of drivers worked for multiple service platforms or covered multiple service categories.

• Advantages Compared to Existing Solutions:

  1. Significant improvement in data generalizability and transparency, covering a global scope rather than being limited to specific regions or single services.
  2. Lowered usage barriers for non-technical research groups through APIs and graphical tools.
  3. Enabled large-scale and high-frequency data collection, enhancing the timeliness and reliability of research findings.

• Experimental or Evaluation Results:

  • Data collected by RMS revealed the significant impact of the pandemic on ridesharing services and uncovered misconceptions in prior supply-demand analyses.
  • For instance, real-time data validated that surge pricing policies typically redistribute existing driver supply rather than significantly increasing total supply.

• Limitations and Future Directions:

  • Limitations:
    1. Data collection is limited to specific small areas within economic centers.
    2. Utilization data may represent upper limits, lacking comprehensive real demand data.
    3. Current focus is primarily on economy-class vehicles, without a comprehensive comparison of ridesharing vehicle models.
    4. Shared server resources are limited, with a data update frequency of 15 minutes per cycle.
  • Future Directions:
    1. Implement cross-category and cross-regional measurements in broader urban areas.
    2. Analyze deeper impacts of multi-factors (e.g., weather, holidays) on ridesharing usage.
    3. Collaborate with HCI researchers to optimize design and evaluation for low-income and disabled groups.
    4. Increase hardware resources to improve data update efficiency and coverage.

Conclusion

This study significantly lowered the barriers to researching and analyzing ridesharing services through the development and application of RMS. It uncovered numerous new findings related to the pandemic and dynamic surge pricing impacts. The research demonstrated the diverse application potential of RMS and provided a valuable prototype and platform for future studies.