Uncategorized

AI Driving Developments Being Made Locally at Arizona State University

It sounds like something Marty McFly would have had in Back to the Future. A car that drives itself…that is something that in the 1980’s and 1990’s seemed like crazy, super optimistic talk. The fact that it is a reality has hit Arizona hard this week, and though the accident in Tempe is a hard lesson in the development of this technology, it is safe to say that artificial intelligence(AI) in driving technology is not going anywhere any time soon.

As a matter of fact, one of the pioneers in driving AI technology works right here at our own Arizona State University. Ashraf Gaffar has been working on artificial intelligence for driving since 2012. Ironically, his resolve to come up with workable technology deepened in 2015 when tragedy struck his own life. His daughter was a pedestrian struck by a distracted driver – the culprit was texting. Though she survived and even exceled at life, it was a long hard road that further inspired and continues to drive Gaffar to remain on the cutting edge of this field.

Gaffar has been working hard to continuously develop artificial intelligence in cars. He is quick to point out that AI is not the same thing as software used in autonomous cars. Software is preloaded scenarios that are limited by their programming. Gaffar prefers AI because it is able to continuously learn and change depending on scenarios. He sees his AI as a helper – almost a “mom” or never-ending drivers ed teacher. Though the AI would come down with pre-taught safe driving procedures, it would be able to not only learn how the individual owner of the car drives, but be able to replicate it, and then run diagnostics on areas where it could be most useful as a safety tool.


Though Gaffar does see autonomous cars as an inevitable breakthrough of the future, he is currently striving to develop driving AI that will serve as that helper more then taking over from the driver immediately. It will be an extra detection system in heavy traffic and be able to take over and get safely off the road if the driver becomes distracted or even disabled somehow. The AI will be able to sense an issue or change in the driver – such as if they fall asleep or have a heart attack. Then the AI can take over, and using its systems safely pull over to the side of the road. These are Gaffar’s immediate goals for his AI system.

Gaffar also made sure to point out that currently most AI and autonomous cars are running in places like here, with pretty consistent, clear climates. With no snow, and little to no fog and rain, it is an ideal weather situation to start the tests. Once you add in places like the Northeast, the issues involved in programming and AI learning increase exponentially. These are also the things that make him favor AI with its ability to learn and adapt to software.

Though cars that fully run themselves are here, they are not without glitches, as we recently and tragically saw. Gaffar and his ASU team will continue to work towards refining the existing technology with greater abilities. With this team on the case, it is clear that the only thing they concern themselves with is learning from mishaps – great and small – and improving the technology so that it can help keep the roads and everyone on and around them safe.