Team places 2nd in IEEE international autonomous racing competition

4 people standing in a classroom.

UB's AudUBon Racing team, from left: Karthik Dantu (faculty advisor), and students Harshita, Rishi Vedula and Harjot Singh Barn.

by Nicole Capozziello

Published December 1, 2021

“AudUBon Racing,” the entry created by a team of computer science and engineering students, was the runner-up in the F1TENTH Autonomous Racing virtual competition. 37 teams from around the world took part in the virtual event.

Print
“In classes, students learn generic approaches to solving problems in autonomy. This competition and others are a great way to apply understanding from classes to the specific conditions of a competition and determine what approaches work best.”
Karthik Dantu, associate professor and team advisor
Department of Computer Science and Engineering

It was the first time that a team from the School of Engineering and Applied Sciences took part in the competition.

Organized by an international community of researchers, engineers, and autonomous systems enthusiasts, the virtual competition took place on Sept. 29, 2021 as part of the IEEE International Conference on Intelligent Robots and Systems (IROS), held in Prague, Czech Republic. 

In the virtual version of the competition, participating teams submitted software that met the race’s simple yet challenging objectives: “don’t crash and minimize laptime.” The competition’s organizers provided teams with rules and guidelines ahead of time, and the virtual track and related virtual infrastructure (server) on the day of.

“It’s exciting for our students to participate in competitions such as the F1TENTH,” says Jinhui Xu, professor and chair of the Department of Computer Science and Engineering. “Testing their approach against others from around the world is an excellent way to grow and demonstrate the capabilities of our department’s students.”

“One of the most valuable things about competitions like these is that they provide an opportunity for goal-oriented work,” says Karthik Dantu, the team’s advisor and an associate professor in the Department of Computer Science and Engineering. “In classes, students learn generic approaches to solving problems in autonomy. This competition and others are a great way to apply understanding from classes to the specific conditions of a competition and determine what approaches work best.”

The team members were Harjot Singh Barn, Rishi Satyanarayan Vedula, and Harshita. Prior to the competition, none of them had a background in robotics. To prepare, Dantu worked with the students over the summer to review relevant literature and study past approaches to the competition.

Award certificate.

When developing their own approach, the team balanced safety with aggression, avoiding the car crashes that hurt some other teams’ performances.

“This is the first time that our team participated in a competition like this.  I enjoyed the atmosphere of the competition the most - meeting every day before and during the competition with the team and trying to do our best for the competition,” said Barn, an undergraduate majoring in computer science and minoring in robotics. “At the same time, this experience helped me apply the skills I have learned in classes to something practical.”

Team AudUBon Racing hopes to someday participate in the in-person competition in which each team brings their own physical F1TENTH car with their software.

Dantu looks forward to continuing to engage students around participating in the competition. In addition, he sees such competition as a means of applying ideas he teaches in his Robotics Algorithms class.