The Path Inference Filter: Model-Based Low-Latency Map Matching of Probe Vehicle Data

We consider the problem of reconstructing vehicle trajectories from sparse sequences of GPS points, for which the sampling interval is between 1 s and 2 min. We introduce a new class of algorithms, which are altogether called the path inference filter (PIF), that maps GPS data in real time, for a va...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on intelligent transportation systems Vol. 15; no. 2; pp. 507 - 529
Main Authors: Hunter, Timothy, Abbeel, Pieter, Bayen, Alexandre
Format: Journal Article
Language:English
Published: New York IEEE 01.04.2014
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Subjects:
Algorithms
Expectation-maximization algorithms
filtering
Global Positioning System
GPS data
Inference
Intelligent transportation systems
Markov random fields
Matching
maximum likelihood estimation
Mobile communication
network-free data
path observation generation
probabilistic map matching
Probability distribution
Probes
Real time
Roads
route choice modeling
Sampling
State of the art
Traffic information
Trajectory
Vehicles
San Francisco California
route choice modeling
GPS data
path observation generation
network-free data
probabilistic map matching
Expectation-maximization algorithms
filtering
maximum likelihood estimation
Markov random fields
ISSN:1524-9050, 1558-0016
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We consider the problem of reconstructing vehicle trajectories from sparse sequences of GPS points, for which the sampling interval is between 1 s and 2 min. We introduce a new class of algorithms, which are altogether called the path inference filter (PIF), that maps GPS data in real time, for a variety of tradeoffs and scenarios and with a high throughput. Numerous prior approaches in map matching can be shown to be special cases of the PIF presented in this paper. We present an efficient procedure for automatically training the filter on new data, with or without ground-truth observations. The framework is evaluated on a large San Francisco taxi data set and is shown to improve upon the current state of the art. This filter also provides insights about driving patterns of drivers. The PIF has been deployed at an industrial scale inside the Mobile Millennium traffic information system, and is used to map fleets of data in San Francisco and Sacramento, CA, USA; Stockholm, Sweden; and Porto, Portugal.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
ISSN:1524-9050
1558-0016
DOI:10.1109/TITS.2013.2282352