- About Us
- Nano-Social Network
- Nano Consulting
- My Account
|Patti D. Hill
CEO / Founder
Penman PR, Inc.
Austin Robot Technology (ART) is an autonomous SUV robot team that is fine-tuning a vehicle capable of driving itself in preparation for the DARPA Urban Challenge in November 2007.
ART's elite team of technologists offers the only Texas entry with a Grand Challenge-tested vehicle. Their Isuzu VehiCross SUV, dubbed "Marvin," was the subject of a course in the department of Computer Science at the University of Texas at Austin, which is a team co-sponsor.
October 3rd, 2007
Marvelous Marvin, the self-driving SUV
Imagine for a moment that you are elderly and can no longer drive. You don't feel well today, and you should see the doctor, but no one is available to drive you there. No worries - just get into your car and tell it where you need to go. A pipe dream? Not any more.
Introducing "Marvin," the self-driving SUV. The product of Austin Robot Technology (ART) and University of Texas at Austin, Marvin uses artificial intelligence and machine-learning technology to literally drive itself. It senses obstacles in its path and makes decisions about how to get around them. It knows the fastest route from point to point. It "sees" and understands street signs and obeys the traffic laws all on its own.
That's better than most teenagers.
Marvin and ART have been chosen by the Defense Advanced Research Projects Agency to advance to the next stage of its Urban Challenge to take place on November 3, 2007. The DARPA Urban Challenge will pit autonomous vehicles from around the globe against one another in a race through an artificial urban landscape in an effort to accelerate research and development in autonomous ground vehicles that could help save lives on the battlefield. The winners will receive $2 million, $1 million or $500,000 and opportunities to commercialize their research.
Some of Marvin's marvelous technology includes:
One of the fundamental problems to be solved as apart of the Urban Challenge is the interpretation and integration of high-density sensory data toward a complete, real-time model of the world that includes an awareness of local terrain and any obstacles or other vehicles in the vicinity. ART is integrating sensory information from several SICK lasers, a GPS sensor, an inertial navigation unit, a state-of-the-art high-density 3D laser sensor, and several video cameras.
Another fundamental problem to be solved as a part of the Urban Challenge is how to achieve precise, safe vehicle control that takes into account this world model, including reaction to observed and predicted behaviors of other vehicles. The team is using modern path planning and localization algorithms to meet this challenge.
Use of machine-learning algorithms, including those for sensor calibration and integration, increases the vehicle's ability to respond to unpredictable environmental scenarios.
Private industry applications of Marvin's technology could include:
ART's innovations in the areas of graphic chip and digital sensor development will allow software industries to offer improved microcontrollers, 3D room mapping, tracking sensors and computer graphics.
Automotive Industry and Automotive Electronics Manufacturing
ART's autonomous vehicle incorporates technologies that could allow the automotive industry to develop valuable systems for use in everyday vehicles, allowing for such options as auto-drive controls and sensors that improve safety by making corrections based on environmental factors, e.g. slows vehicle when road is wet.
ART'S innovative vision system running on graphics processing units could lead to technologies such as mobile vision headsets that could allow the visually impaired to "see."
So how long until millions of Marvins start taking to the road - driving the blind to work, taking kids to soccer practice, taking care of our aging parents? World-renowned artificial intelligence expert Dr. Peter Stone, UT-Austin professor and ART team member expects fully functional autonomous vehicles to become a reality within 10 years.