Flint, MI—After a four-year competition aimed at studying, designing and testing self-driving car technology, a team of Kettering students walked away with multiple awards.
The competition, AutoDrive, pitted eight teams against each other as they each took a donated Chevrolet Bolt and worked to make it fully autonomous. After four years of competition, Kettering’s Bulldog Bolt team placed in fifth place overall out of eight teams and tied for first in one category, the level four automation challenge.
According to the National Highway Traffic Safety Administration and the SAE, there are six levels of automation for vehicles. These levels span from zero, meaning no automation to five, full automation. In between these levels are different driving assist features like lane detection or automatic parallel parking.
Team members at Kettering ended up sharing their first place success for the level four automation challenge with Texas A&M and beating out Michigan State University, Michigan Technological University, North Carolina A&T University, University of Toronto, University of Waterloo and the Virginia Polytechnic Institute and State University.
For context, Tesla’s Autopilot, perhaps the most popular autonomous vehicle system in the world, is only at a level two or partial automation. This means the car’s systems have automated functions including accelerating, braking and steering but still require the driver to be engaged at all times with both hands on the wheel.
Alex Garrow, a graduate student and captain for Kettering’s Bulldog Bolt team, spent much of his time during the competition focusing on the technical aspects of creating an autonomous driving system.
“There are a lot of parts to an autonomous vehicle. One of the most important things is what we call perception … that’s made up of censors such as Lidar (Light Detection and Ranging), radar, and cameras which perceive the world around the vehicle,” Garrow said.
Censors like the ones Garrow described are all designed to work together to create an image of the vehicle’s surroundings that can detect depth, distance, movement, color, and a host of other variables.
The team has all that technology available, but the real challenge is making it all work together.
“Probably the biggest challenge with autonomous vehicles is to be able to detect the world around you to see where the car is, what’s around it, determining from that where it needs to go and what it needs to do. Does it need to stop because there is a deer? Am I in a right turn only and I need to go straight so I have to switch into the other lane?” Garrow said.
Thomas John, project management lead for the Bulldog Bolt team has spent the majority of his time focusing on the logistical side of the challenge.
Aside from science, AutoDrive also requires competitors to show their skills in rigorous bookkeeping and managerial skills.
For John, this meant making sure the Bulldog Bolt team’s several subteams were all meeting their deadlines and staying in touch with each other. It also meant putting together various reports for AutoDrive judges regarding progress made by the time in different sectors of the competition.
“For my side it’s totally different, it’s not based on autonomous vehicles but it’s more about each and every subteam. AutoDrive has many subteams, they have their own way of managing stuff … as project management lead, it’s our job to communicate with all the teams,” he said.
Overall, Kettering took fifth place out of eight, taking 3rd place in the MathWorks Simulation Challenge and the Technical and Safety Reports as well as 4th place in the Social Responsibility Event.
Though June marked the end of the four-year competition, Kettering’s Bulldog Bolt team will be joining the nine-team AutoDrive II competition starting in the fall.
For Garrow and John, who have completed their course work at Kettering, being able to work on cutting-edge technology while learning about team management has set them up for success, they said.
Garrow will start work at the systems integration division of General Motors on Monday and John is currently in talks with a consulting company.
For the team’s faculty advisor, Associate Professor of Mechanical Engineering, Diane Peters, being able to advise a team of students on a project like AutoDrive is rewarding in more than one way. On top of being part of an award-winning team, Peters also looks forward to the way autonomous vehicles will allow individuals with different disabilities to get around on their own terms.
“Autonomous vehicles have a lot of other potentials. I’m epileptic. It’s completely controlled but there are many many people … who are epileptic, who are legally blind, who have various other disabilities and are very limited in their mobility since they can’t drive. So they either have to rely on others or live in an area where there is public transportation and that affects their lives,” Peters said.
The possibility of helping people in the types of situations she described is a common motivation for members of the team.
“If we can perfect autonomous vehicles and make them safe, they can have a huge benefit to society by giving independence and mobility to people with a wide range of limitations,” Peters said.