Page 221 - Big Data Analytics for Connected Vehicles and Smart Cities

202 Big Data Analytics for Connected Vehicles and Smart Cities Practical Applications and Concepts for Transportation Data Analytics 203

determine that the morning peak occurs from 6:00 a.m. to 9:00 a.m. and in
the evening from 3:00 p.m. to 7:00 p.m. Using this definition, the peak periods
span about 27% of the available data points.
Figure 10.3 shows the data for the fourth quarter, as the fourth quarter
was the only quarter for which a direct comparison could be made between
quarter four, 2013, without VSLs and quarter four, 2014, with the VSLs imple-
mented. The VSL system was implemented in September 2014.
Figure 10.3 illustrates the speed variability for the entire study on week-
days during peak periods. This encompasses eight separate quarters, or 520
weekdays of data. The shading indicates the average speeds, and the width of
the line represents again the standard deviation of the traffic speeds. Figure 10.3
illustrates that traffic conditions are in fact highly variable on the freeway, even
during some of the most heavily traveled time periods.
Figure 10.4 shows the speed profile that was assumed for a bottleneck.
This represents a formal data definition for a subjective traffic engineering
term: bottleneck.
Figure 10.4 is adapted from industry best practice with respect to the
definition of a bottleneck in terms of speed [2]. The bottleneck begins when the
observed speed of traffic falls below 60% of the reference speed. The reference
speed is an INRIX [1] data parameter, which represents the 85th percentile of
the observed speeds. This roughly equates to the speed limit enforced on fixed
speed limit signs along freeways. The slowdown is then assumed to persist until
the traffic begins to climb back through 60% of reference speed. Note that this
simplified template describing a bottleneck assumes that there will be only one
slowdown per bottleneck and that when the traffic climbs back through 60% of
reference speed, the bottleneck is over. There is the potential to define a chain
of bottlenecks with one reduction in speed, when, in reality, the chain of bottle-
necks could be one single bottleneck. This will be discussed later in this section.
Figure 10.4 illustrates speed variability across a length of freeway that
is split into a series of TMC sections. A TMC segment is an industry stan-
dard definition of a section of freeway. Note that this should not be confused
with the more general meaning of TMC as a traffic management center. This
definition goes back to an earlier European project known as the Radio Data
System—Traffic Message Channel project (RDS—TMC). Segment definitions
were developed especially for this project to enable location-specific traffic mes-
sages, and they have been adopted on a wider basis since the project was com-
pleted. INRIX [1] adopted TMC segments as the basis for the current version
of the data that was used for this work.
The vertical axis of the graph in Figure 10.4 represents traffic speed in
miles per hour, averaged over one-minute time increments. The horizontal axis
represents distance, with the arrow indicating the direction of travel for the
traffic.