|
|
Mn/DOT TDA Data Automation Project:
The Minnesota Department of Transportation (Mn/DOT) has
been responsible for collecting, analyzing, and publishing traffic count
data from the various roadway systems throughout the state. The traffic
reporting system mainly developed by the Traffic Forecasting and Analysis
Section (TFAS) of Mn/DOT has been used in several federal programs,
internal Mn/DOT applications, and many private sectors. The objective of
this project is to continue the TFAS’ automation efforts by computerized
integration of the current manual efforts to import, filter, and analyze
the TMC portion of traffic data contributed to the Mn/DOT’s Traffic
Monitoring System. All data processing will be performed at the TDRL Data
Center, from which the final outputs will be delivered to Mn/DOT through
on-line access. Additional efforts have been made to deal with missing
data from the raw traffic data by employing statistical data modeling and
multiple imputation techniques.
|
|
|
|
Study On CDF and HDF Archival of ITS
Data
Typical ITS sensor networks cover large geographical
areas such as the area of a whole state and produce data 24 hours a day,
7 days a week, and year after year accumulating a huge amount of data
over the years. As a result, archiving, reusing and sharing of the ITS
collected data became more of a technical challenge. To energize research
and development activities in this area, ITS ADUS program began to
address and promote “archiving and sharing of ITS data to improve
transportation decisions.” Based on these needs, TDRL has developed a
large scaled data model based on archiving of data and management of the
data through a data-center concept. This part of project focuses on
developing efficient archiving technologies, which serves as the basic
building blocks of the data model.
In order to use the archives as the basic building blocks
of the overall data and computational model, we require the archives to
satisfy several properties. They are:
1. The size of the archive must be small;
2. The archive must be portable among
different OS or computer types;
3. Data retrieval of the archive must be
efficient in terms of retrieval time and random access ability;
4. Initial investment and maintenance cost
of the archive must be low;
5. The archive must have a capability of
metadata (description of data),
6. It must follow an open-standard.
We found that these properties are best satisfied by two
data models, (Common Data Format) CDF and Hierarchical Data Format (HDF),
developed by National Space Center Data Center. The first part of this
project is in identifying advantages and disadvantages of using CDF and
HDF for ITS sensor data along with a comparative study on which model
would work better. In the second part, we will be developing various
tools and applications for creation, retrieval, and visualization of CDF
or HDF archive files.
|
|
|
|
ITS Sensor Data Analysis with Missing
Data
All ITS sensors essentially work based on some form of
electronics and mechanical parts to produce data. The sensor data are
then collected to a central location through one or more of communication
links. Because electronics and mechanical parts can never work
permanently without any failure, data loss from sensors is a high
probability. Moreover, road constructions, severe weather, natural
disasters, power outages, and routine maintenances could lead to loss of
data even for the healthy sensors. A serious data loss can occur, if
central severs that collect data from sensors experience a failure or a
malfunction. Therefore, it is safe to assume that ITS sensor data will
always have missing data problems. This research focuses on developing
analysis techniques for the ITS sensor data that contains missing
portions. The main efforts have been directed towards multiple
imputations based on data modeling, such that analysts can perform
regular statistical analysis using the imputed data. Present efforts are
only limited to studying traffic data, but the efforts will be extended
to Mn/DOT’s RWIS data.
|
|
|
|
Design and Management of Modularized,
Efficient, and Secure Transportation Data Center
With the advances in wireless cell phones and Internet
communications, information distribution has become an integral part of
general transportation system. Moreover, data management has become
extremely important as transportation systems are increasingly monitored
through large-scaled ITS sensor networks and computerized reporting
systems. Since the ITS sensor data or reporting systems are sampled and
recorded in real-time, they cannot be reproduced in case of any data
loss. As a result, safely storing and protecting data is of a prime
importance. This project studies how large-scaled ITS data can be managed
in a secure and efficient manner in a large organization. The solution
TDRL arrived at is managing the data through a well-designed data center
(DC) within the state DOTs. The main components of study include
fire-walls, NAS, SAN, back-up strategies, UPS, VPN, servers, and
networking. The TDRL DC will serve as a prototype for this project, from
which various hardware/software options and design alternatives will be
considered.
|
|
|
|
