Augmenting decisions of taxi drivers through reinforcement learning for improving revenues
Taxis (which include cars working with car aggregation systems such as Uber, Grab, Lyft etc.) have become a critical component in the urban transportation. While most research and applications in the context of taxis have focused on improving performance from a customer perspective, in this paper,we...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | text |
| Language: | English |
| Published: |
Institutional Knowledge at Singapore Management University
2017
|
| Subjects: | |
| Online Access: | https://ink.library.smu.edu.sg/sis_research/3867 https://ink.library.smu.edu.sg/context/sis_research/article/4869/viewcontent/15746_68951_1_PB.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Singapore Management University |
| Language: | English |
| Summary: | Taxis (which include cars working with car aggregation systems such as Uber, Grab, Lyft etc.) have become a critical component in the urban transportation. While most research and applications in the context of taxis have focused on improving performance from a customer perspective, in this paper,we focus on improving performance from a taxi driver perspective. Higher revenues for taxi drivers can help bring more drivers into the system thereby improving availability for customers in dense urban cities.Typically, when there is no customer on board, taxi driverswill cruise around to find customers either directly (on thestreet) or indirectly (due to a request from a nearby customeron phone or on aggregation systems). For such cruising taxis,we develop a Reinforcement Learning (RL) based system tolearn from real trajectory logs of drivers to advise them onthe right locations to find customers which maximize theirrevenue. There are multiple translational challenges involvedin building this RL system based on real data, such as annotatingthe activities (e.g., roaming, going to a taxi stand, etc.)observed in trajectory logs, identifying the right features fora state, action space and evaluating against real driver performanceobserved in the dataset. We also provide a dynamicabstraction mechanism to improve the basic learning mechanism.Finally, we provide a thorough evaluation on a realworld data set from a developed Asian city and demonstratethat an RL based system can provide significant benefits tothe drivers. |
|---|