Paving the Road to Success: Hannah Mason

Summer internships are great opportunities for students to have the ability to test drive the classroom knowledge in prospective career fields. But one Lipscomb junior engineering student got a chance this summer to not only test drive her knowledge, but to test drive self-driving cars as well, (at least in computer simulations.) 

 

Hannah Mason, an electrical and computer engineering major from San Diego, California, participated in a research project with the Cognitive and Autonomous Testing Vehicle Program at the University of Arizona. This National Science Foundation funded program offers research experiences for undergraduate engineering students, although it is usually only offered to senior engineering majors. 

 

“They gave us a Ford Escape to work with and they had rewired everything so it didn’t have a manual gas pedal, it didn’t have manual steering–everything was wired so you could control it on a computer,” said Mason. “We were given some software that can simulate the CAT vehicle, so we could test our programs without having to actually test the vehicles.”

 

The CAT Vehicle Program split the students into three groups. Mason’s group researched “a comprehensive lane detection and following system in autonomous vehicles.” After researching what was already accomplished in the field, the group determined how research groups with similar resources detected and followed lanes.

 

 

“It was a lot of sorting the image out by hue, picking out the yellow hues, then sorting the image by brightness and picking out the brightest parts in the image, which usually ended up being the yellow and white lines. Most of my part was refining that process so that it only picked up the lanes (lines.)”

 

In her opinion, Mason believes autonomous cars aren’t quite ready to be on their own. The greatest struggle her research group found was getting the computer to recognize that the sand and the yellow and white cars on the road were not lane lines. 

 

“The sand on the side of the road in Arizona is really bright. We ended up using a perspective transform with a region of interest, which transforms the image from what the driver would see into what a drone overhead would see,” said Mason. “That’s easier to translate into the space of where the car is so the car knows where it is in relation to the lanes.”

 

Although testing of autonomous cars, as the researchers call them, has been going on for some time across the nation, concern about their safety ramped up this spring. There have been a few car accidents involving autonomous cars, but Mason is confident that if autonomous cars ever become mainstream, they will go through rigorous safety protocols. 

 

In an effort to limit accidents and increase the safety of autonomous cars, some researchers are using artificial intelligence to better predict the behavior of pedestrians, Mason explained.

These researchers asked test groups to sort video data of pedestrians into two groups: pedestrians that are about to cross the road and those that are not. The two groups of data sets are logged into the computers as a reference for predicting pedestrian behavior. 

 

Even though she tries to keep her career options open to leave room for God to intercede, Mason said she would love to go to graduate school to research more on this subject, especially in the realm of artificial intelligence. This opportunity gave Mason great experience to benefit her in the classroom and on for her future career. 

 

“I have put this experience on my resume, and it's good to be able to talk about applications of automation and image processing with some real-life context,” said Mason. “One of the best parts of this research experience for undergraduates was getting to work side by side with computer science students. They introduced me to a lot of new computing ideas, which will stick with me through the rest of my studies.”