Adian Cook standing in front of GROVER autonomous bus
Adian Cook wrote the software that allowed GROVER, ORNL’s first autonomous bus, to follow a controller path from point to point. The bus has a novel design, allowing it to spin like a teacup. (hi-res image)

December 18, 2018

Arriving at Oak Ridge National Laboratory in late 2017 as a vehicle system engineer, Adian Cook could not foresee then that he would soon be the example of what happens to a research career when one is in the right place at the right time. But perfect timing is exactly what transpired for Cook as mere weeks into the job, he was presented with the opportunity to work on what he calls a dream project: an autonomous bus.

“I was hired to develop software in the Vehicle Systems Research Group,” Cook said. “This was right when GROVER was put on a rapid pace development schedule and the work needed to make it happen was software.”

GROVER (Ground-based Robotic Omnidirectional Vehicle for Electric-mobility Research) is an autonomous or self-driving, 3D-printed bus that operates on a platform designed to make all four wheels operate independently of each other. Software is the backbone of any vehicle and Cook brought the right mix of coding skills to figure out how to implement GROVER. “With software, you have the framework but there are a million ways to do it. You have all the tools in your toolbox, but then you have to decide what you’re going to deploy to tackle the project,” he said. “We had a huge learning curve with GROVER.”

And with little time to learn, too. GROVER’s development was placed on an accelerated timeframe. Cook and a fellow researcher had to demonstrate the autonomous capability by the beginning of 2018, providing just a few months for software development. In order to begin work, the duo quickly determined that they would first need to test the software on a portable device, which turned out to be a simple utility cart.

“We put the software together and mounted the controller on the cart, along with the GPS and some other equipment,” Cook said. “We called this cart MOUS (Mobile Omni-directional Utility Shuttle).

Cook was soon able to begin testing his code to see if it would guide where the cart needed to move.  “This was an intimidating process. For many weeks it was just the two of us sitting at the board trying to figure out how this works and we had to go from our minds to the whiteboard to MOUS. And we had to have faith, a lot of faith, in our code.”

Feedback from the cart’s controller based on code that Cook developed indicated whether to turn left or right, proving that same code could guide GROVER.

“People saw us in cold weather pushing a cart around at Hardin Valley looking like crazy people, but we were using a $40 cart to test very expensive hardware,” Cook said. “GROVER follows a digital path, so we knew if the cart worked, it would work on the bus. When we ported over to the bus, we had confidence. And it worked fairly well the first time it was tested.

“We’ve had a few hiccups along the way. For example, GROVER object detection is based on LIDAR (a radar system using laser light), and we found that glare from buildings or other cars on sunny days can cause it to stop prematurely. Then, sometimes it would spin in the opposite direction of where it was calibrated because of the GPS waypoint drift,” he continued. “So, we would then rework the code to eliminate the problem. It’s all part of working through the research phase.”

Cook said the next milestone to meet for GROVER includes mounting radars with sensor fusion, which means combining data from multiple inputs and calculating accurate positions and orientation. Looking further into the future, Cook expects machine learning to also be integrated into GROVER.  

From the field to the lab

While Cook has made his mark at ORNL with the autonomous bus, he’s moved onto other projects in ORNL’s Vehicle Systems Integration Lab, recently developing software for a parallel hybrid vehicle to improve fuel economy and reduce costs for an industry partner. Being able to move from one diverse project to another is why working at ORNL is the perfect fit for Cook. 

“I worked in private industry before coming to the lab and in that type of job you don’t get to pick what you do,” he said. “At ORNL, you have the opportunity to look around and discover what you are interested in.”

A first-generation college graduate from Cleveland, Tennessee, Cook didn’t start out with vehicle software programming in mind as a future career option. He originally had his sights on playing football. “I had a 4.0 in high school and a high ACT score, but I always thought of myself as a football player first who just happened to be smart. So, I went to school at Sewanee The University of the South to play football,” he said. “I wasn’t that athletic, but I wanted to play ball and I was good enough to go to division three football and that’s why I chose Sewanee.” 

Cook’s classroom ability quickly surpassed the ball field and four years later, he graduated with a degree in physics and moved on to the University of Tennessee, where he says, the blocks fell into place for him to find his way to ORNL.

He became familiar with ORNL while completing his master’s in mechanical engineering at the University of Tennessee, Knoxville. Then, as a student team member of the EcoCAR project—which challenges students to retrofit conventional cars with hybrid technology to make them more ecofriendly—Cook met ORNL’s group leader for vehicle systems research. 

“Our team retrofitted a Camaro, and this was the first time I had ever been in charge of controls and wiring and software for a vehicle and I loved it” he said. “I was encouraged to consider research work at ORNL when I graduated, and I finally did at just the right time.”

Cook hopes to continue his education in the future too, possibly pursuing his PhD. 

“You know, I work on computers all day, so I ride and race motorcycles to break up the monotony of life,” he added. “But I enjoy engineering—being in the mix of everything and solving technical problems. That’s what keeps me growing and learning.”

ORNL is managed by UT-Battelle for the Department of Energy's Office of Science, the single largest supporter of basic research in the physical sciences in the United States. DOE’s Office of Science is working to address some of the most pressing challenges of our time. For more information, please visit https://science.energy.gov.  

Share