Page 144 - Big Data Analytics for Connected Vehicles and Smart Cities

124 Big Data Analytics for Connected Vehicles and Smart Cities What Are Analytics? 125

“Mean time before failure” is a parameter often quoted with regard to opera-
tional performance of intelligent transportation system devices. This provides
a measure of a single aspect of asset performance and a single data element.
An analytics approach would produce a wider view of asset performance by
taking into account multiple factors such as cost versus performance index,
cost of maintenance, and overall design life in addition to the meantime before
failure. This index would take account of partial as well as total failures. The
index would be used to identify poorly performing devices and as input to an
overall maintenance and replacement strategy; a smart city would also define
and agree to set performance targets and maintenance standards for critical
assets. Analytics can be identified to compare actual maintenance standards
against those implemented and to generate an index that compares the total
cost of maintenance for each device compared to the value delivered by the
device. This approach can extend to all assets within a smart city including
devices, telecommunications assets, vehicle assets, and all other assets associ-
ated with the delivery of the 16 transportation services identified for smart city
implementation. The ability to identify trends and patterns with respect to asset
performance and expenditure on asset maintenance will be a powerful element
in the smart city due to the ability to ensure that resources are being utilized
effectively and that value for money is being achieved. It is to be expected that
life-cycle analytics will also reveal that the cheapest assets (low initial capital
investment) may not present the best value for money over the life of the asset.

Connected Vehicle
It is anticipated that the connected vehicle will provide a richer stream of data
in a big data set that features volume and velocity—with respect to data, that is,
not the vehicle. This will allow the identification use of analytics that make com-
parisons between different data elements. For example, usage-based insurance
professionals would be very interested to measure the number of lane changes
made by a vehicle for every mile. This combined with steering angle compared
the road geometry could provide some valuable insight into driver behavior and
consequent risk. The number of brake applications and accelerator depressions
per mile would form the basis for a driving turbulence index that could form
the basis for crash prediction and certainly provide valuable information on the
effectiveness of traffic signal timings and the number of minutes taken for each
trip across the city; the reliability or variability of each trip would provide valu-
able input into overall transportation performance through the use of trip time
and trip time reliability indexes.

Connected, Involved Citizens
The notion of a connected and involved citizen implies the ability to have a
two-way dialogue with the citizen. In one direction, crowdsourcing, movement