March/April 2004
The AIRS Approach to Analyzing Intersection Crashes
by Jessica Rich
A transportation management center in Kentucky pioneers
a new recording system to improve driver safety
and accountability.
The Federal Highway Administration (FHWA) is dedicated to
achieving success in several aspects of highway management,
but its chief interest always has been to maintain safety on the Nation's
highways. The goal is to reduce roadway-related fatalities and injuries by
designing a forgiving infrastructure and educating road users about
highway safety and technologies. The U.S. Department of
Transportation's (USDOT) goal is to achieve a 33 percent reduction in the motor
vehicle fatality rate by 2008.
|
(Above) An aerial view of the high-crash intersection of Brook
Street and Jefferson Street in downtown Louisville, KY.
|
FHWA has identified intersection safety as one high-risk area of
highway safety. Each year, more than 2.8 million intersection-related
crashes occur in the United States, representing more than 44 percent of all
reported crashes.
The American Association of State Highway and Transportation
Officials (AASHTO) targets intersection safety as one of the emphasis areas in
its Strategic Highway Safety Plan. In November 2001, AASHTO held
a workshop to identify various issues related to intersection safety and
to develop a national agenda on intersection safety. Those participating
in the workshop identified the lack of research focused on the
intersection problem as one important issue and offered a number of concerns
related to research, such as the lack of reliable data on the effectiveness
of safety countermeasures. In addition, participants pointed to the need
for improved understanding of intelligent transportation systems
(ITS) and data on failure mechanisms in the driver decisionmaking process.
To address these issues, AASHTO has been conducting research
and evaluating advanced technologies for intersection safety. "Researchers
and engineers need to work together to discover the problems at
intersections," says Barney Leslie, a project manager for Traffic Response
and Incident Management Assisting the River Cities (TRIMARC) in
Louisville, KY and southern Indiana. "The
Auto Incident Recording System (AIRS) is a cooperatively produced tool
that will aid in this effort."
AIRS is a sound-actuated video recording system used to help
analyze the reasons for traffic conflicts at intersections. The idea for
AIRS came about when Makoto Koura, of the Mitsubishi Electric
Company, invented a monitor/recorder to analyze and improve his golf swing
and then later applied the idea to development of AIRS by the
Japanese company in 1995. The system was first implemented in Japan to
help traffic police analyze high-incident intersections, study the causes
of these crashes, and implement countermeasures such as installation
of pavement markings to help prevent future incidents.
About AIRS
AIRS consists of two video cameras located on two corners of the
intersection to capture incidents from different perspectives. The
system also includes two directional microphones (one per camera) that
listen for "crash-like" sounds such as "metal-against-metal" contact,
broken glass, horns honking, and squealing brakes. A series of digital signal
processors and recording media transmit the sounds and images to
a video cassette recorder (VCR) for easy access and analysis.
The video cameras capture three stages of each crash or
"near-incident" before, during, and
immediately after. Video and sounds are recorded continuously on an
8-second digital memory loop. This continuous loop eliminates the task
of watching countless hours of video that do not contain crashes.
When the system detects an evident collision, another 4 seconds of
video capture ensues. Then the 8-second loop is transmitted to a video
recorder in the unit, capturing the sights and sounds of the
intersection 4 seconds prior to the crash, during the instant of the crash, and 4
seconds after the crash. Afterwards, the system returns to recording
8-second loops until another crash or "near-miss" incident occurs. The
system also captures the phase of the traffic signals by monitoring the current
to the signal head and pedestrian signs. The signal phase then is
encoded onto the recorded video for review. A benefit of the system is that
near-incidents triggered by tires squealing or horns honking allow the
traffic engineer even greater insight into traffic patterns at the
intersection and driver habits.
The cameras can be mounted easily onto existing signal poles
or onto temporary wood poles. The interface to the signal system
is nonintrusive and therefore does not require any modification or
merging with the intersection signal controller. The installation of the
entire system takes about 2 days. In addition to the video cameras and
directional microphones, the installation package also includes a cabinet
that houses the AIRS controller, its digital video and signal processors, a
VCR and its controller, and a signal phase detection unit.
The system also contains incident analysis software, a simulation
program that makes it possible to view each phase of the incident, the
point of contact, and the speed of the vehicles during the incident.
Traffic officials, including crash reconstructionists, police, and city traffic
engineers, use these pieces of information when analyzing the incident
to determine problem areas and potential improvements to the
intersection. Traffic officials need a VCR to view the videos, and they typically use
a scanner in conjunction with the incident analysis software.
Setting Cooperative Goals
In 2001, TRIMARC became the first transportation management group
in North America to test AIRS. Serving as the transportation
management center for the greater Louisville, Kentucky/Southern Indiana
urbanized area, TRIMARC is committed to using intelligent transportation
systems and innovative transportation technologies to save lives, time,
and money. TRIMARC posed several specific goals to pursue in
conjunction with AIRS. The team plans to work with the Kentucky
Transportation Cabinet to achieve a 10 percent decrease in intersection crashes
by fiscal year 2008. By collecting data at a high-incident intersection
within the city and studying the frequency and types of incidents occurring
at the intersection or approaching it, TRIMARC hopes to lay the
groundwork for achieving this goal.
The second key goal is to reduce the crash average at one
of Louisville's most dangerous intersections. TRIMARC has been using
AIRS for about 2 years at the high-crash intersection of Brook and
Jefferson streets in downtown Louisville. The intersection includes two
one-way streets with adjoining exit ramps from I–65 (southbound). A gas
station is located at the southwest corner and a hotel at the northwest
corner. Major hospitals in Louisville are located within six blocks of the
intersection, and a fire station is less than a block away, contributing to
Brook and Jefferson's average of one crash per week (283 in a 5-year period).
|
A video camera mounted on a temporary wood pole shows
one way to install AIRS.
|
The partners hope to reduce the crash rate at this intersection
by studying driver behavior and using this information to determine
countermeasures. Local traffic engineers implemented, monitored, and
modified countermeasures at this intersection using information
drawn from AIRS to analyze their effectiveness. TRIMARC's goal is to
deploy the system just long enough to gather sufficient data
to increase safety at this intersection and then move the
system to another high-incident location. The system also
may be deployed eventually to intersections without traffic signals.
Effective Countermeasures At Brook-Jefferson
Engineers' review of the AIRS data over the first 5 months found
that turns from the Brook Street ramp were the predominant cause
of crashes at the intersection and were occurring at a rate of 1.25
per month. The city of Louisville made several modifications to the
intersection and evaluated the effectiveness of those modifications, using the
AIRS video. Changes included the extension of the traffic island for the
Brook Street ramp traffic and additional signage on lane assignments for
motorists going north on Brook Street.
During the 6 months following these modifications, only
three crashes involving vehicles from the Brook Street ramp occurred.
Since then, the intersection has experienced only 15 crashes in almost
2 years. Previously, the crash rate was nearly 5 crashes per million
vehicles entering the intersection. Now, the rate is 0.70 crashes per million
vehicles. In addition, incidents associated specifically with the turns
from the Brook Street ramp have been reduced to a current rate of .67
per month, a 50 percent reduction from the previous rate of such
crashes. "AIRS thus has proven itself already
as a reliable analytical tool," says Leslie.
|
The AIRS analytical tool can retrieve a frame from each portion of a single incident. |
AIRS allows for almost instantaneous analysis of the effectiveness
of countermeasures because of its ability to use real-time data for a
detailed study of the recorded "near-miss" events without having to
wait for relatively infrequent crash events. Benefits include improved
planning for future intersections from analyzing the mistakes in current
designs and reduced crash rates that, in turn, will decrease fatalities, injuries,
and damage to property. Reducing the incidence of crashes and
injuries also will lessen the associated monetary costs.
Perhaps the most significant benefit of AIRS is its portability.
Once the system has been used to improve one intersection, it can
be moved to another high-crash location. This feature contributes to
the system's economy and effectiveness.
The Future of AIRS
The quick analysis capability provided by AIRS has been highly
effective in improving the traffic efficiency and safety at the Brook
and Jefferson intersection, especially by providing thorough
documentation of crash events occurring during the study period. Since the
Brook and Jefferson deployment is the first AIRS installation in the
United States, engineers have monitored the intersection closely for nearly
2 years, evaluating the effectiveness of the system and gathering
traffic information.
|
A vector can be drawn from each car and linked
to determine the angle of collision. This
information can help in analyzing errors in the design
of the intersection.
|
Leslie believes that the 2-year history of the effectiveness and value
of AIRS is now sufficient for other transportation and safety engineers to
take advantage of the system. "I really expect to start to see deployments
to other areas of the country," says Leslie.
He also believes that the video has given city and State traffic
engineers additional insight into the dynamics of the intersection and
the causes of incidents and near-incidents. Previously, that
information was not readily available or was too costly to acquire since it would
have required 24/7 surveillance.
The Kentucky Transportation Center (KTC) at the University
of Kentucky College of Engineering in Lexington plans to conduct its
own evaluation of the overall effectiveness of AIRS. The center will
evaluate both the cost-benefit ratio and past success as a data collection
and analysis tool. The final goal of this study is the development of a
recommendation concerning whether additional systems are
warranted and, if appropriate, a plan for deployment. The intent of the study
is to provide the Kentucky Transportation Cabinet (KYTC) with an
understanding of the relative gains associated with detailed data collection
of intersection incidents and incident recording systems.
The results of the study will determine whether the Cabinet
purchases and installs another system at another high-incident
intersection. The Kentucky Transportation Center anticipates that the final
evaluation and report will be completed by June 2005.
Award Winner
In 2002, ITS America recognized TRIMARC and KYTC with the
"Best of ITS" award in the area of
research. In October 2003, KYTC was picked from 128 applicants to receive
the FHWA Office of Safety's National Roadway Safety Award in the
category of program planning, development, and evaluation, for its work
in using AIRS. The deployment of AIRS was a successful partnering
effort between the Kentucky Transportation Cabinet, Northrop
Grumman Mission Systems, the Louisville Metro Police Department, the Metro
Department of Public Works, and the Mitsubishi Electric Company.
"I think everyone assumed a little risk in being the first in the
country to attempt this, but I believe the safety results we are now
attaining justifies everyone's willingness to see this through," says Leslie.
The distributor continues to work with other communities,
engineering firms, and the insurance industry to encourage the use of the AIRS
at high-incident intersections.
Jessica Rich is in her first year as a highway safety engineer for
the FHWA Kentucky Division. Rich has a B.S. degree in civil and
environmental engineering from North Carolina Agricultural and Technical
(A&T) State University and an M.S. in civil engineering from the University
of Tennessee. Rich has previously worked for the North Carolina
Department of Transportation, the Transportation Institute of
North Carolina A&T State University, the Center for Transportation
Research at the University of Tennessee, the city of Knoxville, and the
Knoxville Transportation Planning Organization. Her past experiences
include surveying, research, transportation planning, and traffic engineering.
The author would like to thank Daniel Woo of Northrop
Grumman Mission Systems, Jerry Pigman of the Kentucky Transportation Center,
and John Crossfield of the Kentucky Transportation Cabinet for their input
in preparing this article.
Other Articles in this issue:
Hyperfix 65/70
Coordinating Incident Response
Erosion Control with Recycled Materials
Glenwood Canyon 12 Years Later
A Tale of Two Canyons
Spotlight on the South
The AIRS Approach to Analyzing Intersection Crashes
Resource Center Goes National