Principal Investigator

Sean Qian

Status

Completed

Start Date

July 1, 2017

End Date

June 30, 2018

Project Type

Research Advanced

Grant Program

FAST Act - Mobility National (2016 - 2022)

Grant Cycle

2017 Mobility21 UTC

Visibility

Public

Abstract

Central to smart cities is the complex nature of interrelationships among various urban systems. Linking all urban systems is the system users. The individual daily activities engages using those urban systems at certain time of day and locations. There may exist clear spatial and temporal correlations among usage patterns of all urban systems. The objective of this research is to fuse and analyze massive data from transportation, energy, and social media systems to discover the spatio-temporal correlations of usage patterns among those systems.  Two questions will be addressed using the data collected in the City of Pittsburgh and CMU buildings energy use. 1) What can we tell about the morning commute by knowing households’ utility or social media use the night before? and how can we optimally manage the morning commute using this new information?  2) What can we tell about the evening commute by knowing building energy use or social media activities during the daytime? and how can we optimally manage the evening commute using this new information?    

Description

Central to smart cities is the complex nature of interrelationships among various urban systems. Linking all urban systems is the system users. The individual daily activities engages using those urban systems at certain time of day and locations. There may exist clear spatial and temporal correlations among usage patterns of all urban systems. The objective of this research is to fuse and analyze massive data from transportation, energy, and social media systems to discover the spatio-temporal correlations of usage patterns among those systems. This enables cross-system demand prediction and management. Transportation system can be used minutes or hours ahead of energy system(s) and/or social media activities as a result of daily activity chains. Therefore, the spatio-temporal usage of the transportation system can be accurately predicted a few minutes or hours ahead by real-time sensing user patterns of other urban system(s). This is otherwise hard to accomplish by solely monitoring a transportation system. Ultimately, real-time control strategies for demand management of the transportation system can be developed with efficient real-time demand prediction.

This research will be focusing on two specific subproblems. 1. What can we tell about the morning commute by knowing households’ utility or social media use the night before? and how can we optimally manage the morning commute using this new information? The congestion starting time and duration of roadway, bus and parking can vary by 30-60 mins from day to day, but this information is central to morning-commute transportation management. Unfortunately, they are difficult to predict by solely sensing traffic/passenger flow due to demand randomness. The free-flow traffic/passenger flow in the early morning does not exhibit clear patterns before it transitions to being congested (also known as traffic “break-down”). When integrating
transportation systems into other urban systems, the occurrence time of congestion
may be partially linked to commuters’ activities the night before, or hours ahead in the
early morning. Using data mining models, we can reveal the utility use patterns for households, and identify their potential impacts to the morning commute. The congestion starting time and duration for all modes of transportation in the morning commute may therefore be predicted based upon their spatio-temporal relationship with energy use and social media activities. This information enables efficient transportation management strategies to optimally reduce congestion and emissions. Examples of strategies include, but are not limited to, providing information to travelers, offering incentives for individuals, and optimal pricing of transportation systems. 2. What can we tell about the evening commute by knowing building energy use or social media activities during the daytime? and how can we optimally manage the evening commute using this new information? Similar to the subproblem 1, our
hypotheses is that individuals’ departure time for evening commute can be partially
revealed by their daytime activities in the office buildings and/or social media activities. Therefore, energy/social media use at around noon-4pm may help us understand individuals’ activity patterns and further help predict evening commute congestion. Such spatio-temporal correlations can be mined using historical measurements of utility usage, building occupancy and traffic/passenger flow.

TECHNICAL PROPOSAL FORM

PI Qian directs the Mobility Data Analytics Center of CMU since 2013. The center has established an integrative transportation data platform that has the full capacity to archive, manage and analyze massive data collected from multi-modal multi-jurisdictional sources. In particular, the following data sets in the Allegheny County are available to accomplish this research, • GIS data for buildings (parcels), parking locations near CMU, all roads and bus transit systems. • Real-time traffic speed data covering major highway and arterials, including all main roads in the vicinity of CMU • Traffic counts covering some highway segments and some arterials including some main roads in the vicinity of CMU • Three years of time-varying parking occupancy data for on-street parking and publicly-owned garages in the vicinity of CMU • Two years of time-varying parking occupancy data for some garages on CMU campus
• Three years of public transit passenger counts and vehicle location data for all bus/light rail trips and stops • CMU building energy use data at the 1 min level for the last two years • Place-tagged and geo-tagged Twitter data in the City of Pittsburgh for the last two year

Timeline

2 years

Strategic Description / RD&T

Deployment Plan

Work with PennDOT and the City of Pittsburgh

Expected Outcomes/Impacts

1. A methodology estimates the travel time on main roads using hours-ahead energy data and/or social media data.   2. An open-source web application implementing the methodology. The tools, programs and data will be made available in the public domain.

Expected Outputs

TRID

Individuals Involved

Email	Name	Affiliation	Role	Position
xpi@andrew.cmu.edu	Pi, Xidong	CEE	Other	Student - PhD
seanqian@cmu.edu	Qian, Sean	CEE	PI	Faculty - Research/Systems
shuguany@cmu.edu	Yang, Shuguan	CEE	Other	Student - PhD

Budget

Amount of UTC Funds Awarded

$95000.00

Total Project Budget (from all funding sources)

$95000.00

Documents

Type	Name	Uploaded
Publication	User-centric interdependent urban systems: using time-of-day electricity usage data to predict morning roadway congestion	March 25, 2018, 11:25 a.m.
Presentation	Mobility data analytics	March 25, 2018, 11:25 a.m.
Presentation	Mobility data analytics	March 25, 2018, 11:25 a.m.
Progress Report	34_Progress_Report_2018-03-30	March 25, 2018, 11:25 a.m.
Progress Report	34_Progress_Report_2018-09-30	Sept. 23, 2018, 7:20 p.m.
Final Report	34_-_user-centric-report.pdf	Oct. 22, 2018, 4:51 a.m.
Publication	From Twitter to Traffic Predictor: Next-Day Morning Traffic Prediction Using Social Media Data.	Dec. 2, 2020, 9:37 a.m.

Match Sources

No match sources!

Partners

No partners!