INTELLIGENT TRANSPORTATION SYSTEMS DATA

WAREHOUSES AND THEIR APPLICATIONS

Yunjie Zhao

, Madhubabu Sandara

, Shan Huang

, Adel Sadek

Department of Civil, Structural, and Environmental Engineering, University at Buffalo, SUNY, Buffalo, NY, U.S.A.

Department of Computer Science and Engineering, University at Buffalo, SUNY, Buffalo, NY, U.S.A.

Tom George, Athena M. Hutchins

Niagara International Transportation Technology Coalition (NITTEC), NY, U.S.A.

Keywords: ADMS, ITS data warehouse, Applications.

Abstract: Archived Data Management Systems (ADMS) offers an opportunity to take full advantage of the data

collected by Intelligent Transportation Systems (ITS) devices in improving transportation operations and

planning at a minimal additional cost. This paper develops an ITS data warehouse, or an ADMS, which can

be used to a wide range of applications, such as more effective transportation planning, decision-making and

performance measurement, extreme traffic event study, traveller information subscriber system, data

support for model development, calibration and validation and so forth. In short, the development of such a

data warehouse is going to be beneficial for both traffic management and research purposes.

1 INTRODUCTION

On a worldwide scale, transportation accounts for

about 21% of CO

emissions, with surface

transportation representing the largest source

accounting for more than 90% of the CO

emissions

produced from all transportation modes (Gorham,

2002; EEA, 2003).

Among the strategies being explored for

improving traffic conditions, and hence reducing

energy consumption and harmful emissions, are

those falling under the umbrella of Intelligent

Transportation Systems (ITS) technologies. The

basic philosophy behind ITS is to take advantage of

recent advances in information technology,

communications, and advanced computing to

improve efficiency, safety and environmental

compatibility. The focus of this paper is on one

specific ITS application, namely Archived Data

Management Systems (ADMS).

Our data warehouse mainly focuses on the

Niagara Frontier Corridor, the border region that

encompasses the Niagara River border crossings. In

Western New York, it represents a strategic

international gateway or corridor of critical

importance to trade and tourism flow between the

United States and Canada. According to the

Canadian Consulate General of Buffalo,

approximately 30% of the total Canada-US trade

crosses at the Niagara border, along with millions of

immigrants and tourists every year. A report by the

Ontario Chamber of Commerce (OCC) in 2005 puts

the value of the annual land-borne merchandise

crossing the Buffalo-Niagara Frontier border at

$60.3 billion dollars (OCC, 2005).

An archived data warehouse is an invaluable

asset for the transportation systems in Niagara

Frontier corridor, not only for research purposes, but

also for traffic management. It supports several

applications for improving mobility, sustaining

economic development, reducing fuel consumption

and minimizing emissions.

Archived Data Management Systems (ADMS)

offer an opportunity of take full advantage of the

travel-related data collected by Intelligent

Transportation Systems (ITS) devices in improving

transportation operations, planning and decision-

making at a minimal additional cost. ADMS are

designed to archive, fuse, organize and analyze ITS

data and can therefore support a wide range of multi-

layer applications. Examples of such applications

include:

343

Zhao Y., Sandara M., Huang S., Sadek A., George T. and M. Hutchins A..

INTELLIGENT TRANSPORTATION SYSTEMS DATA WAREHOUSES AND THEIR APPLICATIONS.

DOI: 10.5220/0003431503430347

In Proceedings of the 13th International Conference on Enterprise Information Systems (ICEIS-2011), pages 343-347

ISBN: 978-989-8425-53-9

 2011 SCITEPRESS (Science and Technology Publications, Lda.)

(1) Developing more effective operational

strategies (e.g. optimizing traffic signal);

(2) Planning for operations and special events

(e.g. inclement weather and snow storms);

(3) Enhancing traveller information systems by

providing the added capability of forecasting

future conditions, e.g. border crossing delay

estimation;

(4) Long-term planning and transportation

investment decision-making;

(5) Performance measurement benchmarking and

reporting.

Given this, an ITS data warehouse is a key

component of any integrated corridor management

(ICM) approach.

2 LITERATURE REVIEW

The idea of developing ADMS or ITS data

warehouse have been officially proposed since 1996

(ADUS Program, 2000; Liu et al, 2002), though the

concept itself has been existing for even longer time.

Initially, research studies were conducted, aiming at

developing standards and guidelines for developing

such system. These studies resulted in suggesting

best practices for: (1) data archiving and fusion; (2)

data screening and imputation techniques; (3) data

modelling and mining methods; (4) archived data

functions and applications; and (5) data presentation

and dissemination techniques, among many other

aspects (FHWA, 2009).

Following these research initiatives, several

states in the United States took leading roles in

implementing ITS data warehouses. For instance, in

Maryland, one major objective of the University of

Maryland’s Center for Advanced Transportation

Technology Laboratory (CATT Lab) is to serve as

an AMDS system to meet different data needs from

national, state, and local levels (CATT Lab, 2009).

In Kentucky, an ADMS has been developed to

archive and disseminate the data collected by two

ITS deployment, i.e. ARTIMIS and TRIMARC

(Chen and Xia, 2007).

With respect to the development and applications

of ADMS or ITS data warehousing systems, the

State of New York has lagged behind other states

around the country.

For example, in the Greater Buffalo‐Niagara

area, key traffic data such as traffic counts, travel

time, accidents, and border crossing delays are

separately collected and maintained by different

transportation organizations, and no formal data

integration and archiving mechanism currently exists

due to institutional, technical or budget barriers. This

situation hinders the full realization of the utility of

ITS data for transportation operations, management

and evaluation.

3 ITS DATA WAREHOUSE

3.1 Architecture

Figure 1: Data warehouse structure.

Figure 1 above shows an overview of the ITS data

warehouse for the Niagara Frontier Corridor. As can

be seen, the data warehouse is envisioned to serve as

a data repository for a wide range of useful

traffic‐related data streams. Those data are

transferred, processed and archived in the data

warehouse, where the end users are able to perform

a one-stop data query to retrieve the data they are

interested in. So far, the ITS data warehouse system

has archived or is planning to archive the following

data and data sources soon:

 Traffic volume on major arterial and freeway

with hourly interval;

 Incident data maintained by NITTEC, namely

Incident log and help log which contain

information about the incident location, start

time, end time and so forth;

 Weather data (e.g. temperature, Hourly

precipitation, visibility, etc.);

 Turning movement counts at major

intersections;

 Travel Time Data from the TRANSMIT

ICEIS 2011 - 13th International Conference on Enterprise Information Systems

344

system, which uses roadside readers to

identify EZ pass or toll tags;

 Border crossing delay (delay time of US-

Canada border);

3.2 Star Schema

Most of the data are stored as the classic STAR

Schema. This is because the primary referencing

system used within the data warehouse is link based,

which means that most of the data, except some

node–based data like intersection turning

movements, could be related to a certain link. For

instance, when an accident happens, the incident log

and help log would record its location and this

location is joined to the specific link in the data

warehouse as Figure 2 shows.

Figure 2: Star Schema for incident log.

3.3 Data Import

Since the data warehouse is archiving a wide range

of data resources, each using a different kind of

format (e.g. XML, spreadsheets and so forth), a

standard procedure for importing and archiving all

of these data streams into the database was needed.

The import process was designed to

automatically extract raw data into the archived

format on a regular basis. For example, NITTEC

would dump the incident XML files onto the server

everyday via a scheduled script; our program detects

the new files, process them and archive all the

information into the data warehouse. All of these

processes can be done via the scheduled batch files.

3.4 User Interface

The user interface of the data warehouse connects

the database and the end user. So far, a simple web

interface, implemented using GeoServer in order to

provide for a geographic map for users to select

graphically the route segment they are interested in,

if provided. (Figure 3) The interface supports the

following services: historical data query, traffic data

on selected roads (Travel times, volume), real-time

travel time and incident mapping and so forth. Other

functionality will be developed according to the

demand of the users. End users could click the query

routes they are interested in, and select the specific

type of data they are looking for as well as favourite

time slot.

Figure 3: the user interface window.

4 APPLICATIONS

4.1 Transportation System

Performance Measurement &

Management

First of all, a comprehensive network performance

measurement system could be developed based on

the travel time, border delay and other traffic data

archived in the data warehouse. In other words,

transportation agencies like NITTEC could use the

measurement system to better communicate the

“state” of the transportation system to the public and

policy makers. Moreover, these performance

measures indices can be used to identify potential

issues in the current system, and thus help the

transportation agencies to develop effective

operational and management strategies to improve

mobility and reliability of the transportation system.

Secondly, traffic signal optimization could be

conducted with the data archived in the systems.

Several recent studies of signal optimization and

coordination show reductions in the number of

vehicle stops, as a result, ranging from 6 to 74%,

with the magnitude varying depending upon the

congestion level (Sunkari, 2004). Studies also show

that reducing the number of vehicle stops could have

significant positive environmental impacts.

In addition, the archived data could be mined to

gain useful insight into the transportation system

performance and its problems. For example, a report

could be generated to provide a summary of the

frequency and time slots of the incidents on a certain

INTELLIGENT TRANSPORTATION SYSTEMS DATA WAREHOUSES AND THEIR APPLICATIONS

345

routes, and to study whether there is any common

factor among them. Besides, one could study the

impact that incidents have on the transportation

system performance, and how long it typically takes

to clear incidents, and bring operations back to

normal.

Other than the applications aforementioned, we

are currently performing a research study to develop

predictive models for border crossing delays.

4.2 Extreme Traffic Events

The Greater Buffalo Niagara region is well known

for its winter weather which is characterized by

numerous and sometimes severe ‘lake-effect’ snow

storms. These events result in significant delay and

increase the frequency of accidents. The data stored

within the data warehouse can provide an

opportunity to better understand the impact of such

events on the transportation network, and hence can

help in devising emergency plans for dealing with

such disrupting events.

The Transportation Analysis and Simulation

System (TRANSIMS) is an integrated, open-source

set of transportation planning models designed to

provide a number of capabilities that go beyond the

traditional “four-step” modelling process. The

TRANSIMS framework has four components: a

population synthesizer, an activity generator, a route

planner, and a micro-simulator. Also, the area has

recently been selected as one among a handful sites

nationwide for the test deployment of the

TRANSIMS model, focusing on freight border-

crossing issues.

On Dec 2nd, 2010, for example, a severe snow

storm hit the south Buffalo area, which forced the

New York State Thruway Authority to shut down

the Thruway (I-90) for several hours.

The data stored can thus be used to study how

traffic flow behaviour and patterns change during

such events, and then to devise effective

management strategies for dealing with such

situations. We are currently in the process of

correlating the archived weather data to TRANSMIT

travel time information to perhaps develop models

that show us how travel time or traffic speed

changes with the different weather and road

conditions

4.3 Traveller Information System

In a related effort, we developed a system called

MYNITTEC, which is a personalized subscriber

traveller information system that allows users to

receive customized real-time traveller information in

Western New York and Southern Ontario via text

messaging and/or email. Subscribers have the ability

to select specific expressways, days of the week and

times of day to correspond with your travels.

The way MYNITTEC works is simple. Users

choose their favourite routes and time spots as their

unique travel profiles. Each travel profile allows the

user to receive personalized notifications from the

system. The data warehouse could help enhance the

traveller information system by providing both

predictive information and more detailed route

performance measurement. On the other hand,

MYNITTEC collects some useful information about

travellers in the Western New York region (e.g. their

preferred routes, times of travel). This information

will also be stored within the data warehouse and

can be mined to understand certain aspects of travel

behaviour in the region.

4.4 Model Development, Calibration

& Validation

4.4.1 Model Development

Aside from the traffic data archived in the data

warehouse (i.e. traffic volumes, travel times,

accidents, etc.), the data warehouse also includes

very useful static information about the attributes of

the transportation network. For example, the system

has stored very detailed link attributes, like the

number of lanes, pocket lane, length of the links and

so forth, which provides an opportunity to automate

and facilitate the process of developing traffic

simulation models to support the different traffic

studies in the region.

As we know, there are all kinds of simulation

models, such as AIMSUN, CORSIM, PARAMICS,

VISSIM and so forth. Although some of them are

macroscopic model, some are mesoscopic or

microscopic model. One thing they have in common

is that they all require a network to build the model

on, although they all have their own network file

format. Therefore, another direct application of data

warehouse is the model development based on the

network information we’ve already archived. In

other words, if we are able to transfer the network

information into the format a simulation software

could read, it could make model development a lot

easier instead of an extremely time consuming

process especially for large-scale networks.

ICEIS 2011 - 13th International Conference on Enterprise Information Systems

346

4.4.2 Model Calibration & Validation

After a simulation model is built, large volumes of

data are required to calibrate and validate the

simulation results. In that regard, the ITS data

warehouse plays a significant role as a data provider.

We are actually currently doing exactly this in a

research project to develop and calibrate a large-

scale TRANSIMS model, an agent-based model

originally developed by Los Alamos National Lab,

of the Buffalo-Niagara region. Within the subarea

we are modelling in TRANSIMS Micro-simulator in

TRANSIMS, there are 193 count stations. The

following example is to just give a simple idea how

the data warehouse could benefit us. We

summarized the 24-hr trip distribution based on all

of those count stations between the model and

reality (Figure 4).

After the calibration

Figure 4: Trip Distribution of TRANSIMS model and field

counts after the calibration.

5 CONCLUSIONS

The archived ITS data warehouse for the Niagara

Frontier Corridor supports a wide range of

applications designed to improve mobility, sustain

economic development and reduce fuel consumption

and emissions. The benefits associated with the

different applications of the data are detailed below:

 The data archived in the system support the

development of a comprehensive

performance measurement framework for the

area, especially for the international border

crossings, and their associated delays. The

data warehouse could also support signal

optimization, which not only improves the

transportation system, but also reduces the

tail-pipe emission as well as fuel consumed ;

 The data warehouse provides traffic data and

weather information to better understand the

impact of the extreme events like inclement

weather on the traffic and driving behaviour;

 The data can also enhance traveller

information systems in the region by adding a

predictive component to the traffic data

provided;

 The warehouse can also support the

development and calibration of traffic

simulation models; and The Niagara Frontier

ITS data warehouse could serve as a model

deployment for other regions around the State

of New York, which would benefit from the

lessons learnt from this study.

REFERENCES

CATT Lab (2009). CATT Laboratory, A Research Center

at the University of Maryland, College Park. http://

www.cattlab.umd.edu/index.php?page=about. Last

access: September 15, 2009.

Chen, M., and J. Xia (2007) “Archived Data Management

System in Kentucky.” Kentucky Transportation

Center; Kentucky Transportation Cabinet; and Federal

Highway Administration. 39p.

Data mining, http://en.wikipedia.org/wiki/Data_mining

Eng-Wong Taub Associates, MRC, Bergmann Associates,

and TranSystems (2009) NITTEC Transportation

Operations: System Operational Concept, Draft

Technical Memorandum #6. 50p.

European Environment Agency (EEA). (2003).

Greenhouse gas emission trends and projections in

Europe 2003, Tracking progress by the EU and

acceding and candidate countries towards achieving

their Kyoto Protocol targets. Environmental issue

report no. 36, ISBN 92-9167-632-2, EEA,

Copenhagen.

Federal Highway Administration (FHWA) (2009).

Archived Data User Service (ADUS) Program

Activities. http://www.fhwa.dot.gov/policy/ohpi/tra

vel/adus.cfm. Last access: September 15, 2009.

Gorham, R. (2002). Air pollution from ground

transportation an assessment of causes, strategies and

tactics, and proposed actions for the international

community. The Global Initiative on Transport

Emissions. A Partnership of the United Nations and

the World Bank, United Nations.

Liu, H., R. He, Y. Tao, and B. Ran (2002). A Literature

and Best Practices Scan: ITS Data Management and

Archiving. Wisconsin Department of Transportation.

Sunkari, Srinivasa. “The Benefits of Retiming Traffic

Signals,” ITE Journal. April 2004. Costs ID: 2007-001

INTELLIGENT TRANSPORTATION SYSTEMS DATA WAREHOUSES AND THEIR APPLICATIONS

347