Seminar Series: Technology Trends 2004
Smart Cars: The next frontier for ICT technologies
Alex Zelinsky, CSIRO ICT Centre
Information and Communication Technologies are now beginning to
have a major impact in the automotive industry. Increasingly, the
basic functions in motor vehicles are becoming computer controlled,
and telecommunications is being used to add new functionality to
vehicles. A major challenge for automobile makers is to make vehicles
safe. In Australia over 1,700 people die on the roads every year,
while worldwide over 750,000 people are killed by automotive accidents.
The motor vehicle is one major cause of death in modern societies.
Intelligent vehicles could make a significant impact in making our
roads safer and reducing the road toll.
It is envisaged that eventually the automobile will evolve into
an autonomous robotic system where all the functions in the vehicle
are computer controlled. Electronic sensors and motors will replace
the hydraulic and mechanical systems of a vehicle. Imaging sensors
will monitor the external environment and will allow the automobile
to automatically control behaviours such as advanced cruise control,
lane keeping and collision avoidance. In-car sensors will allow
monitoring of the driver for safety systems to guard against
fatigue, distraction and inattention. Drivers will be able to
interact and control vehicle functions through natural interfaces
such as speech and visual gestures. Communications technology will
open up an enormous range of automotive applications. The latest
traffic and environment conditions will be provided to drivers and
in-car navigation systems, in-car entertainment systems will have
access to audio/video resources and the internet, in the event of an
accident, breakdown or an emergency the exact position and status of
the vehicle can be transmitted. The market for applications of
wireless telecommunications in motor vehicles or "telematics"
is predicted to explode, with a market size of $US13 billion within
10 years[1].
Seeing Machines is developing smart technology for cars that will
contribute to safer road traffic in the near future. One of the
company's first developments is a system that provides driver
assistance support. A pair of video cameras aimed at the driver and
a computer that analyses the images to determine the position and
orientation of the driver's head and the direction in which they are
looking. The immediate application of this system is to provide a
warning if the driver starts to go to sleep or is inattentive.
Seeing Machines is also studying other applications of the robotic
technology to provide safer cars. This includes combining the video
cameras observing the driver with cameras observing the road (vision
in and out of vehicles). Using the driver monitoring system in
combination with data from cameras observing the road, it would be
possible to warn the driver if there is an obstacle on the road that
they haven't seen. For example, if the driver starts to change lanes
without checking, and the external cameras detected a car or other
obstacles in the way, a warning is given. The company is also
developing computer vision technology for road scene understanding,
including surrounding vehicle tracking, pedestrian detection and
road sign recognition.
The seminar will describe the real-time computer vision
algorithms that have been developed for Smart Cars which operate
robustly in highly dynamic situations. Demonstrations by video of
the various computer vision systems will be provided.
Seeing Machines
Seeing Machines (see www.seeingmachines.com
) was founded in 2000 as a spin-off from the Australian National
University and Volvo to commercialise computer vision technology for
the automotive field. The company is the world leader in its field,
employing 22 staff with revenues in excess of $3 million a year.
faceLAB is a Seeing Machines' product that is designed to assist in
the study of human behaviour by tracking where a person is looking.
The technology consists of a pair of video cameras, a PC-workstation
and the software to perform the vision processing. The software
detects features on a person's face, such as eyes, corners of the
mouth and eyebrows, and tracks the position of these to estimate the
position and orientation of the head. In addition, the pupils are
tracked to determine where the person is looking (the gaze
direction). The measurements on head position, gaze direction and
blinking are calculated at 60 frames per second. The system is
non-intrusive, needing no headgear, markers or other physically
attached equipment. It may be used in simulator, laboratory or
in-vehicle settings, day or night.
Seeing Machines products are being used within the automotive
industry by Volvo, Bosch, Delphi, Renault, Motorola, Nissan, Honda,
Mitsubishi, Daimler Chrysler and Toyota. The company is partnering
with Volvo to develop a driver safety product for Heavy Trucks.
Seeing Machines has won numerous national and international awards
for its technology & business achievements.
[1] Intelligent Transport Systems America www.itsa.org
Back to "Technology Trends"
|
