194 PART | V The future of ITS applications
human-operated cars. Among the services that can be delivered over the con-
nected car technology is the ability to warn for blind spots, for sudden braking of
the leading vehicle or forthcoming obstacles, or even to notify for approaching
emergency vehicles. Wi-Fi-based short-range communications (e.g., dedicated
short-range communication) seemed to be the first option, but existing cellular
standards (e.g., 4G or LTE) are currently gaining popularity and become prom-
ising with the advent of 5G standard. Data transfer speed, response times, and
data security are the main concerns for companies that develop connected car
solutions and regulators.
Artificial Intelligence is another enabler of self-driving vehicle technol-
ogy. The abundance of sensor data and the increasing ability of car-embedded
systems to perform complex computations moves the interest of researchers
from “simple” machine learning algorithms to deep artificial neural network
architectures that enable the in-vehicle systems to learn and make its decisions.
Learning by the input of a fleet of cars instead of individual instances gives to
deep and reinforcement learning models an advantage towards other solutions
(KPMG, 2016 ). The application of deep learning techniques to continuously
collected real-life data from different traffic situations allows autonomous cars
to become better and better in analyzing and understanding their context and
reacting in an informed manner. The future for autonomous systems of that
level of expertise seems to be accident free and safe for drivers and pedestrians.
Apart from the in-vehicle technology enablers, there are several IT and IoT
technology advancements in the network and infrastructure that allow autono-
mous cars to take informed decisions (Johnson, 2017). The road network can
be an essential part of the ITS, by hosting a network of on-road devices, which
accurately detect road and traffic conditions and early warn approaching vehi-
cles or even activate actuators spread over the road network (e.g., smart traffic
lights). Communication with the vehicles is supported by a range of roadside
communication devices such as beacons located alongside the road or in the
place of traffic signals and communication with the road network operating
center can be achieved by connecting the sensors and actuators to the city fiber-
optic network. This limits solutions to the urban environment, whereas for high-
ways wireless technology can be employed. The replacement of traffic signals
by communication devices that transmit information to the nearby vehicles is a
step that further improves autonomous driving, but still has many open issues to
be solved. For example, the bad weather conditions that interfere with sensors
and introduce noise or interrupt their signal (Ng & Lin, 2016). The redundancy
of sensors and sensor types can be a solution to this problem.
17.2.1 Beyond technology enablers
Apart from the technology enablers that promote the concept of autonomous
driving, there exist more things that have to be considered, as we are moving
to higher levels of autonomy. For example, the expectations of the society for