Technology Profile |
Measuring pedestrian safety
using automated conflict analysis
The intersection of Spring
Street and 5th Street in
midtown Atlanta, Georgia
experiences a large volume of
pedestrians and bicycles. The
observed conflicts led to the
opportunity for installing an
all-walk pedestrian phase, also
known as a pedestrian scramble.
Unfortunately, pedestrians
and other vulnerable road users
(VRUs) are disproportionately
impacted in terms of road traffic
fatalities. In the US, they
accounted for 48% of all road
traffic fatalities in 2016.1 An
intersection’s safety is typically
measured by the number of
crashes that occur at the
location. However, pedestrian
crash data is typically under
reported, therefore crash reports
often lack sufficient detail to
determine appropriate action.
An effective pedestrian
safety evaluation technique to
effectively quantify pedestrian
safety before and after the
integration of an all-walk
pedestrian phase into the signal
cycle at the intersection was
seen as a necessary tool in
the project.
In January 2018, a pedestrian
scramble phase was integrated
into the signal cycle at the
intersection. The project
deployment team; including
Midtown Alliance, as well as
engineering firms Jacobs and
Aecom, decided to use Brisk
Synergies’ on-demand
automated video-based road
conflict analysis solution.
Temporary cameras were
installed, and BriskLumina
analysis was used to track and
classify all individual road users
from video data, to
automatically detect near-misses
and other surrogate safety
indicators that are indicative
of future potential collisions.
Video data was collected
before and after the
Traffic Technology International November/December 2019
www.TrafficTechnologyToday.com
046
Above: The all-walk
phase temporarily turns
the intersection into
a car-free zone
implementation of the all-walk
phase with the potential conflicts
of drivers and pedestrians at the
intersection evaluated.
The all-walk phasing not
only increases pedestrian
visibility, but also adds spatial
separation by placing
pedestrians in the center of the
road, so chances for collisions
are reduced. This type of signal
phasing also increases the
efficiency of the intersection
for pedestrians by allowing
them to cross diagonally, and
therefore complete their
crossing movement during one
signal phase instead of two.
Methodology
For conflict analysis, the main
surrogate safety measures used
to determine conflict risk and
identify an improvement in
pedestrian safety from before
to after are Post-Encroachment
Time (PET) and vehicle speed.
PET is used to calculate the
proximity of each conflict. Six
scenarios were analyzed for 6
hours each during the before
and after conditions.
The video footage was
collected on-site and
automatically processed using
machine learning and artificial
intelligence. It is important to
note that schools were in session
and weather conditions were
clear (no rain or snow) during
the video collection phase.
Results
The results indicated that the allwalk
pedestrian phase provided
a significant benefit to
pedestrians. Overall, the
percentage of pedestrians
involved in near collisions (PET
< 3 seconds) at the intersection
reduced from 4% to 1% after
the intervention, representing
a 75% improvement.
The solution provided
trajectory heat-maps in order
to visualize road user paths. At
this particular intersection the
turning vehicle speed increased
from 5 to 8 mph for rightturning
vehicles, and from 12 to
16 mph for left-turning vehicles.
Using Brisk Synergies video
analytics tools, Aecom was able
to have an accurate
measurement of that speed
increase. According to Mark
Start, a senior ITS traffic
engineer at Aecom, “This is
really a measure of driver
aggressiveness and we felt this
data was helpful to explain
to the lay person. With an
increased speed does come
a greater exposure for a crash
that could hurt somebody.”2
The tool also allowed for
distinguishing between the
vehicle-pedestrian and the
vehicle-bicycle behaviors as
the characteristics were very
different between the two.
This was very useful for
dealing with speed concerns
at the intersection.
| Need to know
The lowdown on
BriskLumina …
> An on-demand,
automated, video-based
road conflict analysis
solution
> Capable of tracking and
classifying all road users
> Detects near misses and
likely causes of future
accidents automatically
/www.TrafficTechnologyToday.com