178 PART | V The future of ITS applications
reported to the driver. The long list of ADAS services comprises—(1) services
that improve safety such as automatic braking for avoiding collisions, blind-
spot detection, automatic lane-keeping, and changing, driver monitoring and
drowsiness detection, (2) services for preventive vehicle maintenance such as
tire pressure and engine monitoring, (3) services that facilitate driving such as
adaptive cruise control, automatic parking, hill-descent control, traffic informa-
tion, intelligent-speed adaptation, and automotive night vision, etc.
Collision avoidance applications
The contribution of machine intelligence to the collision avoidance problem
has a long history and targets all three elements involved in a collision (i.e., the
vehicle, the driver and the environment) and all types of vehicles from cars to
airplanes. Collision prevention systems use long (radars, LIDARs, AIS, etc.)
and short-range sensors for detecting nearby vehicles (blind-spot detection) or
obstacles ahead and trajectory prediction algorithms, and notify the driver or
nearby drivers to avoid risks.
The use of AIS has been enforced in naval transportations, so all cargo ves-
sels (and many more transportation vessels) are obliged to transmit their posi-
tion, direction, etc every few minutes using radio frequencies. The respective
collision avoidance system of a vessel continuously collects AIS data from its
nearby vessels, analyzes their trajectory information, and predicts their near
future position. Based on the predicted future positions and the vessel actual
trajectory, they are able to early inform the vessel captain and avoid a potential
collision. The interest for collision avoidance systems has arisen in the case
of road-transportation networks and urban-transportation networks due to the
advent of (semi-) self-driving vehicles. The same happened for vessels and
unmanned aerial vehicles (Mahjri et al., 2015).
In the case of road-transportation networks, the typical applications that
relate to smart-transportation networks for the benefit of drivers were limited to
a list that included collision avoidance, collaborative cruise control, automatic
detection of driving code violations, emergency alerts (e.g., concerning weather
and road conditions, other vehicles, etc). More recently the ability to operate
pretrained machine learning models that analyze driver’s behavior and vital sig-
nals as well as the vehicle’s condition allowed to act preventively, to predict
potential vehicle malfunctions and early detect driver’s distraction and fatigue
(Mukhtar et al., 2015).
Tesla (Enhanced) Autopilot (https://www.tesla.com/autopilot), is a driver-
assistance system, which uses a camera for lane centering, and changes lanes
when the driver agrees, offers self-parking and adaptive cruise control to and
from the parking spot.
Information mediation
Faster network connections between vehicles and the road-side units of the
transportation network allowed several applications to emerge that enhance the