Purdue, West Point to Partner on Autonomous Race Car

Date
12/11/2020

 PDF

Students and faculty from Purdue University and the U.S. Military Academy at West Point are working together on a team entry for the Indy Autonomous Challenge. (Image provided)

The Indy Autonomous Challenge will bring together collegiate teams from around the world next October to run autonomous race cars around the Indianapolis Motor Speedway’s famed 2.5-mile oval.

To get to the starting line, collegiate teams have been working since 2019 bringing together interested faculty and students, writing business plans and raising funds.

The challenge, set for Oct. 21-23, is a series of events that has teams address topics such as artificial intelligence, sensors, data science and more in the successful operation of a high-speed autonomous race car. Events include themed hackathons, research papers, simulation races, and participation in Purdue University’s autonomous evGrand Prix at IMS.

For the challenge, Purdue is partnered with the U.S. Military Academy at West Point to form Black & Gold Autonomous Racing. Both groups have been working virtually with each other during the pandemic.

“The Black & Gold team is doing very well – both as a team and in finishing the challenges,” said Danny White, manager of Purdue Motorsports and managing director of the evGrand Prix.

“They, as well as all the other teams, are working on issues of how does an autonomous race car pass another race car, speed up, slow down if there is a caution or stop if a race is red-flagged, or even pull into the pits and exit safely to get back onto the track,” White said. “These are really complicated tasks that we take for granted.”

Aly El Gamal, an assistant professor in Purdue’s School of Electrical and Computer Engineering, says the Indy Autonomous Challenge is a great opportunity for students to develop and test state-of-the-art algorithms and address challenges related to navigation at high speeds.

“The technical challenges are very rich in a sense that they involve a broad range of problems across mechanical, electrical and computer engineering fields,” El Gamal said. “The opportunity to compete against top research labs in the world in this challenge is very exciting, as we get to observe the behavior of their algorithms in challenging situations and compare that to the solutions we implemented.”

According to El Gamal, Black & Gold achieved second place in the fastest lap challenge and scored sixth overall at the last hackathon Sept. 12-22. The team can’t rest on its accomplishments as another hackathon is scheduled for Saturday (Dec. 12) through Jan. 22. The hackathon’s competition phase will run Jan. 4-22.

“We are preparing by developing collision avoidance, overtaking, as well as advanced control algorithms, as the next hackathon will involve head-to-head races, so we need to develop intelligent algorithms that can adapt to a wide range of possible competitor strategies,” El Gamal said.

“The students are very excited about this. It’s a special opportunity to build collaborations between student teams, and connect graduate and undergraduate students with similar interests and expertise,” El Gamal said.

El Gamal is advising five graduate students and is working with the more than 30 undergraduate students who are members of Autonomous Motorsports Purdue (AMP) and AMP Vertically Integrated Projects. The students are divided into six subteams developing software, electrical, navigation and mechanical systems. Participating students come from the College of Engineering, Polytechnic Institute, College of Science and other academic units across Purdue.

The team also is preparing for the autonomous section of the evGrand Prix. They will have access to lidar sensors, which help determine objects’ locations, and other technologies to make a go-kart complete a lap around a temporary track at IMS.

“Our undergraduates have also made significant progress toward having a go-kart ready for participation in the evGrand Prix, and hence, we can prepare for the physical race by testing our algorithms on the go-kart first, before deployment on the Indy Lights car,” El Gamal said.

For more information, visit Purdue University.

RELATED

 


-->