Optimizing ride-sharing potential in New York City: a dynamic algorithm analysis of peak and off-peak demand scenarios

This study used the dynamic algorithm to analyze the ride-sharing potential in New York City (NYC) by comparing the no-sharing scenario with different sharing scenarios. At first, the data set was analyzed and categorized based on the characteristics of trips. Then, no-sharing scenarios were defined as a benchmark against which to compare sharing outcomes. Then, to do sharing scenarios, at each step, a specific percentage is subtracted from the estimated number of taxis in the zero scenarios, and various key performance indicators (KPIs) are compared to determine the optimal number of taxis for the demand of NYC during peak hours and off-peak hours from the viewpoints of customers and taxi companies. The study found that the number of needed vehicles can be reduced to 40% for peak hours on weekends and 50% on weekdays.