[Paper Review] OpenStreetCab: Exploiting Taxi Mobility Patterns in New York City to Reduce Commuter Costs

Vsevolod Salnikov, Renaud Lambiotte+2 more|arXiv (Cornell University)|Mar 10, 2015

Human Mobility and Location-Based Analysis2 references21 citations

TL;DR

This paper proposes OpenStreetCab, a mobile application that compares real-time taxi fares between New York City's yellow cabs and Uber X using a publicly available 2013 taxi dataset and Uber's API. It finds that yellow taxis are on average $1.40 cheaper than Uber X, especially for short trips under $35, and demonstrates how Uber's pricing model exploits the prevalence of short-distance travel in urban mobility patterns.

ABSTRACT

The rise of Uber as the global alternative taxi operator has attracted a lot of interest recently. Aside from the media headlines which discuss the new phenomenon, e.g. on how it has disrupted the traditional transportation industry, policy makers, economists, citizens and scientists have engaged in a discussion that is centred around the means to integrate the new generation of the sharing economy services in urban ecosystems. In this work, we aim to shed new light on the discussion, by taking advantage of a publicly available longitudinal dataset that describes the mobility of yellow taxis in New York City. In addition to movement, this data contains information on the fares paid by the taxi customers for each trip. As a result we are given the opportunity to provide a first head to head comparison between the iconic yellow taxi and its modern competitor, Uber, in one of the world's largest metropolitan centres. We identify situations when Uber X, the cheapest version of the Uber taxi service, tends to be more expensive than yellow taxis for the same journey. We also demonstrate how Uber's economic model effectively takes advantage of well known patterns in human movement. Finally, we take our analysis a step further by proposing a new mobile application that compares taxi prices in the city to facilitate traveller's taxi choices, hoping to ultimately to lead to a reduction of commuter costs. Our study provides a case on how big datasets that become public can improve urban services for consumers by offering the opportunity for transparency in economic sectors that lack up to date regulations.

Motivation & Objective

• To compare the cost-effectiveness of yellow taxis and Uber X in New York City using publicly available mobility and fare data.
• To investigate how Uber’s pricing model leverages patterns in human mobility, particularly the high frequency of short trips.
• To develop a practical mobile application that enables commuters to choose the most affordable taxi service based on real-time price comparison.
• To promote transparency in urban transportation pricing by leveraging open data to empower consumers.
• To demonstrate how big data can improve urban services and reduce commuter costs through informed decision-making.

Proposed method

• The study uses a longitudinal dataset of 2013 New York City yellow taxi trips, including pickup/drop-off coordinates and total fares (including tips).
• For each trip, the fare is compared to the estimated fare from Uber’s API for the same origin-destination pair using the Uber X service.
• Uber’s API returns a price range; the mean of the minimum and maximum values is used for comparison.
• A spatial indexing system with 100m×100m grid cells enables efficient lookup of similar trips in the taxi dataset for any user query.
• The OpenStreetCab app uses geolocation to find nearby taxi trips in the dataset and retrieves corresponding Uber X fare estimates via the API.
• The system provides real-time price comparison, with future enhancements planned for time-of-day and traffic-aware pricing.

Experimental results

Research questions

• RQ1 How do the fare prices of New York City yellow taxis compare to those of Uber X for the same origin-destination pairs?
• RQ2 To what extent does Uber’s pricing model exploit the statistical regularity of short-distance urban trips?
• RQ3 Can a mobile application that compares taxi fares in real time reduce commuter costs in urban environments?
• RQ4 How accurate and scalable is a price estimation system based on historical taxi data and real-time API queries?
• RQ5 What role can open data play in increasing transparency and competition in urban mobility services?

Key findings

• Yellow taxis were found to be on average $1.40 cheaper than Uber X for the same trips, based on median fare comparisons.
• Uber X was more expensive than yellow taxis for 70% of trips below $35, indicating a cost disadvantage for shorter journeys.
• The mean trip distance in New York City was 2.09 kilometers, reflecting the dominance of short-distance travel in urban mobility.
• The study confirms that human mobility patterns—characterized by a heavy-tailed distribution of trip distances—favor pricing models like Uber’s that charge higher base fares for short trips.
• The OpenStreetCab prototype successfully demonstrated real-time fare comparison using a combination of historical taxi data and live Uber API queries.
• The application is deployable on mobile platforms and provides a functional tool for commuters to make cost-informed transportation decisions.