Taxi Dispatch With Real-Time Sensing Data in Metropolitan Areas: A Receding Horizon Control Approach

Abstract

Traditional taxi systems in metropolitan areas often suffer from inefficiencies due to uncoordinated actions as system capacity and customer demand change. With the pervasive deployment of networked sensors in modern vehicles, large amounts of information regarding customer demand and system status can be collected in real time. This information provides opportunities to perform various types of control and coordination for large-scale intelligent transportation systems. In this paper, we present a receding horizon control (RHC) framework to dispatch taxis, which incorporates highly spatiotemporally correlated demand/supply models and real-time GPS location and occupancy information. The objectives include matching spatiotemporal ratio between demand and supply for service quality with minimum current and anticipated future taxi idle driving distance. Extensive trace-driven analysis with a data set containing taxi operational records in San Francisco shows that our solution reduces the average total idle distance by 52%, and reduces the supply demand ratio error across the city during one experimental time slot by 45%. Moreover, our RHC framework is compatible with a wide variety of predictive models and optimization problem formulations. This compatibility property allows us to solve robust optimization problems with corresponding demand uncertainty models that provide disruptive event information. Note to Practitioners-With the development of mobile sensor and data processing technology, the competition between traditional "hailed on street" taxi service and "on demand" taxi service has emerged in the US and elsewhere. In addition, large amounts of data sets for taxi operational records provide potential demand information that is valuable for better taxi dispatch systems. Existing taxi dispatch approaches are usually greedy algorithms focus on reducing customer waiting time instead of total idle driving distance of taxis. Our research is motivated by the increasing need for more efficient, real-time taxi dispatch methods that utilize both historical records and real-time sensing information to match the dynamic customer demand. This paper suggests a new receding horizon control (RHC) framework aiming to utilize the predicted demand information when making taxi dispatch decisions, so that passengers at different areas of a city are fairly served and the total idle distance of vacant taxis are reduced. We formulate a multi-objective optimization problem based on the dispatch requirements and practical constraints. The dispatch center updates GPS and occupancy status information of each taxi periodically and solves the computationally tractable optimization problem at each iteration step of the RHC framework. Experiments for a data set of taxi operational records in San Francisco show that the RHC framework in our work can redistribute taxi supply across the whole city while reducing total idle driving distance of vacant taxis. In future research, we plan to design control algorithms for various types of demand ...

Multi-Agent Reinforcement Learning for Order-dispatching via Order-Vehicle Distribution Matching

“…Other automatic order-dispatching methods [12,19] focus on reducing the pick-up distance or waiting time by nding the nearest orders. While these methods usually fail to reach a high success rate on order-dispatching and ignore many potential orders in the waiting list which may be more suitable for vehicles.…”

A Predictive Taxi Dispatching System for Improved User Satisfaction and Taxi Utilization

“…An approach to taxi dispatching with real-time sensing data in metropolitan areas is presented in [9]. In particular, the authors present a framework to dispatch taxis, which combines highly spatio-temporally correlated demand/supply models and real-time GPS location and occupancy information.…”

One-Shot Coordination of Feeder Vehicles for Multi-Modal Transportation

“…Recently, there is also a growing interest in routing problems with awareness of the demand and the fleet. For example, [17], [18] optimally dispatch taxis based on the location of the taxis and customer requests. While many routing problems deal with discrete vehicles and discrete entities to be transported, macroscopic models that deal with flows of vehicles and volumes of demand and supply are also common.…”