Achieving higher taxi outflows from a drop-off lane : A simulation-based study

Abstract

Lanes used by taxis and other shared-ride vehicles at airports and rail terminals are often congested. The present paper examines congestion-mitigating strategies for a special type of lane inside of which taxis are prohibited from overtaking each other while dropping-off patrons. Taxis must therefore often wait in first-in-first-out (FIFO) queues that form in the lane during busy periods. Patrons may be discharged from taxis upon reaching a desired area near the terminal entrance. When wait times grow long, however, some taxis discharge their patrons in advance of that desired area. The Nanjing South Railway Station in China is selected as a case study. Its FIFO drop-off lane is presently managed by police officers who allow taxis to enter the lane in batched fashion. Inefficiencies are observed because curb space near the upstream and downstream ends of the lane often goes unused. A microscopic simulation model is developed in-house, and is painstakingly calibrated to data measured in the study site's FIFO lane. Simulation experiments indicate that rescinding the lane's present batching strategy can increase taxi outflow by more than 25%. Further experiments show that even greater gains can be achieved by requiring taxis to discharge patrons when forced by downstream queues to stop a prescribed distance in advance of a desired drop-off area. Further gains were predicted by requiring the lead taxi in each batch to discharge its patron(s) only after travelling a prescribed distance beyond a desired location. The above findings are confirmed for scenarios calibrated with field data collected on two different days, and for hypothetical scenarios with varying input parameters. Roles that technology can play in implementing these new lane-management strategies are discussed. So are their practical implications in light of the present boom in shared-ride services.