[Paper] Fare Comparison App of Uber, Ola and Rapido

Source: arXiv - 2512.04065v1

Overview

The paper presents a web‑based fare‑comparison app that pulls real‑time pricing from three major Indian ride‑hailing platforms—Uber, Ola, and Rapido—and surfaces the cheapest, fastest option for a user‑specified destination. By stitching together disparate APIs and handling location‑based queries, the authors aim to give commuters transparent, data‑driven choices and improve the overall ride‑hailing experience.

Key Contributions

• Unified API layer that aggregates fare estimates from Uber, Ola, and Rapido into a single response.
• Python‑driven backend that normalizes disparate data formats, performs price‑time trade‑off calculations, and returns the “best” ride.
• End‑to‑end prototype (web UI + backend) demonstrating real‑time fare comparison for arbitrary origin‑destination pairs.
• Practical discussion of integration challenges (rate‑limiting, authentication, emulator‑based testing with Appium, and geolocation handling).
• Open‑source‑ready architecture that can be extended to additional mobility services (e.g., bike‑share, public transit).

Methodology

1. Data Acquisition – The authors reverse‑engineered the public APIs of Uber, Ola, and Rapido (using documented endpoints where available and web‑scraping otherwise). Authentication tokens are refreshed automatically.
2. Location Normalization – User‑entered latitude/longitude pairs are fed into each service’s “price estimate” endpoint. The system reconciles differences in coordinate precision and map‑projection quirks.
3. Fare & ETA Fusion – For each provider, the backend extracts the estimated fare range and expected arrival time (ETA). A simple scoring function (score = α * fare + β * ETA) ranks the options; the weights α and β are configurable.
4. Web Front‑end – A lightweight Flask app renders a form for origin/destination entry, displays a side‑by‑side table of the three services, and highlights the top‑ranked ride.
5. Testing Harness – Android Studio’s emulator and Appium scripts simulate ride‑request flows to verify that the API calls stay in sync with the mobile apps’ behavior.

Results & Findings

• Cost advantage: Rapido (motorbike) is cheapest on average, but its longer ETA can make Uber or Ola preferable for time‑sensitive users.
• Transparency boost: Users who consulted the app saved ≈12 % on fare compared to picking a service at random.
• Technical feasibility: The unified backend handled ~150 concurrent requests with sub‑second latency, proving that real‑time aggregation is practical at modest scale.

Practical Implications

• For developers: The paper’s modular API‑wrapper can be dropped into existing travel‑orchestration platforms, enabling multi‑provider price comparison without building each integration from scratch.
• For product teams: Embedding a fare‑comparison widget directly into a mobility‑as‑a‑service (MaaS) portal can increase user trust and reduce churn by offering “best‑price guarantees.”
• For ride‑hailing companies: Transparent pricing dashboards may pressure providers to tighten their fare‑estimation algorithms, potentially leading to more competitive pricing structures.
• For end‑users: A single click reveals the cheapest ride for a given trip, saving money and reducing decision fatigue—especially valuable in price‑sensitive markets.

Limitations & Future Work

• API stability – The approach relies on third‑party endpoints that can change without notice; a more robust solution would involve formal partnership agreements or use of standardized industry APIs (e.g., Mobility Data Specification).
• Scoring simplicity – The current linear combination of fare and ETA ignores factors like vehicle type, surge pricing, user preferences, and safety ratings. Future work could incorporate machine‑learning models that learn personalized weighting from user behavior.
• Geographic scope – The prototype is limited to a few Indian metros; scaling to nationwide or cross‑border coverage will require handling regional pricing rules and regulatory constraints.
• Real‑time surge handling – The system snapshots fare estimates; integrating continuous streaming updates would improve accuracy during peak demand periods.

By addressing these gaps, the fare‑comparison framework could evolve into a full‑featured, production‑grade service that powers the next generation of transparent, user‑centric mobility platforms.

Authors

• Ashlesha Gopinath Sawant
• Sahil S. Jadhav
• Vidhan R. Jain
• Shriraj S. Jagtap
• Prachi Jadhav
• Soham Jadhav
• Ichha Raina

Paper Information

• arXiv ID: 2512.04065v1
• Categories: cs.LG, cs.AI
• Published: December 3, 2025
• PDF: Download PDF