Page 231 - Big Data Analytics for Connected Vehicles and Smart Cities

212 Big Data Analytics for Connected Vehicles and Smart Cities Practical Applications and Concepts for Transportation Data Analytics 213

Toll Return Index Concept

Toll agencies face a specific problem with respect to justifying the level of tolls
that users must pay. In many cases, the focus is placed on the amount of the toll
to the detriment of a simple explanation of the benefits delivered in return for
the toll. The challenge is that the level of toll paid is a simple number, while the
benefits delivered may be distributed across multiple departments and multiple
budgets within the agency. The toll return index concept attempts to address
this challenge by making use of analytics to generate a simple number that
represents the benefits delivered in return for the toll. The toll return index is a
single number that represents the amount of toll paid divided by the total ben-
efits delivered. The total benefits delivered are comprised of safety, efficiency,
and user experience benefits. The concept was developed in close association
with a leading toll agency and the relevant professional society [3].

Analytics Used
In this example, the toll return index is composed of four primary components:

• Toll paid: There are a number of different ways in which the toll paid
could be characterized. In this example, the total tolls paid by a user dur-
ing an average commute day will be used. This means that the benefits
portion of the toll return index will also be calculated in the same basis
to determine the index.
• Safety benefits: The value of safety benefits is related to the number of
crashes avoided through the increased safety on the toll road. The in-
creased safety can be because of better traffic management, better road
geometry, or better overall operations. It is determined by comparing
crash statistics for the subject freeway to comparable local, regional, and
national roads. Crash statistics are also normalized by considering the
volume of traffic. Therefore, the crash statistics will be quoted in terms
of crashes per vehicle mile traveled. Crash statistics will be obtained
from local crash databases. Traffic volumes may be obtained from traffic
count data available from the local transportation agency, or from the
toll agency itself. In some cases, traffic volumes may be estimated using
mathematical simulation models. Crash statistics are typically classified
as follows:
• Fatal;
• Injury;
• Damage-only.