High school students race the autonomous mini-cars they built this summer at MIT

A select group of students spent the summer learning about the development of self-driving vehicles.
A select group of students spent the summer learning about the development of self-driving vehicles. –Allegra Boverman

Boston.com Cars is your go-to resource for coverage of local car news, events, and reviews. In the market for a car or truck? Check out our new car specials and used car specials curated by our local dealer network.

For some students, summer camp means four weeks immersed in sports or roasting s’mores around a campfire.

But a select group of students from around the country spent the summer at the Massachusetts Institute of Technology learning about the development of self-driving vehicles.

A total of 46 high school students from top STEM schools across the country were invited to participate, with about half the students coming from Greater Boston. Others came from as far away as California, Arkansas and Mississippi.

Advertisement

The program was hosted by MIT’s Beaver Works Summer Institute, a collaboration between MIT’s Lincoln Laboratory and the MIT School of Engineering. Over the course of four weeks, students learned how to develop algorithms and software with the purpose of teaching a miniature car navigation, mapping and object detection.

The program wrapped up on Friday with the students racing their vehicles against each other in MIT’s Walker Memorial. Inside the building, the cars raced around a miniature replica of Boston’s defunct Grand Prix racecourse.

“We see a lot of autonomy slowly making its way into our lives,’’ said Dr. Robert Shin, the program creator. “Autonomous vehicles are a hot topic, but turns out it’s an incredibly interesting research project as well.’’

Parth Parekh, a 16-year-old student from Bergen County Academics in Hackensack, New Jersey, said the program was both “very challenging and at the same time very fun.’’

“This was a very fast-paced course,’’ said Parekh. “We were moving from topic to topic, but it never felt stressful. It never felt like they were rushing us and [it was] really enabling to learn a lot while at the same time not failing under stress.’’

Parekh said teaching robots tasks that come easily to humans, like making the correct turn at a three-way intersection, was the most challenging part of the program.

Advertisement

“The difficult thing is realizing that robots, despite the incredible advances that we’ve made in recent years, are not even close to the intelligence we can display,’’ said Parekh. “And trying to translate what we easily see into the robot through code is, I think, the biggest hurdle [for self-driving cars].’’

Another participant, 17-year-old Sydney Zheng from Thomas Jefferson High School for Science and Technology in Alexandria, Virginia, echoed that assessment.

“The biggest thing that was a challenge was realizing that there are many parts of reality you can’t always address,’’ said Zheng. “The battery might not work, lighting might not be good. You have to consider a lot of different parts of reality that are not easily addressable by well-planned algorithms.’’

Even Shin acknowledges that the material covered in the program can be challenging.

“I couldn’t take this course,’’ he admitted. But he’s hopeful that participating students can carry the problem-solving lessons with them as they embark on their careers.

“It’s really hard to have this many cars drive autonomously and be safe and try to go fast,’’ said Shin. “I think they learned to appreciate what it means to work on a real problem.’’

Loading Comments...