Plowing through code: Robotics student cultivates innovation with agricultural robots
Non-traditional robotics student helps designs agricultural robots
As a long-haul truck driver, Kevin Wing transported newsprint from the Inland Northwest to locations in California, stopping along the way at eateries known to the drivers of the 18-wheelers as the best food joints along the route.
“There was this little mom and pop-type place south of Bend, Oregon, …,” he recalls, but the name of the truck stop, like a lot of aspects of his former occupation have melded into the hazy past.
These days, Wing, a nontraditional student from Post Falls who earned a GED more than 20 years ago and who is a member of U of I’s robotics team on the Coeur d’Alene campus — located at North Idaho College — is helping to build robots and artificial intelligence data systems. Wing is also looking forward to earning a bachelor’s degree in computer science in Spring 2025.
With the help of an Office of Undergraduate Research SURF grant (SURF stands for Summer Undergraduate Research Fellowship), Wing was able to help develop a weeding robot that will be used by the US Forest Service to kill weeds in its seedling beds at the government’s forest nursery in Coeur d’Alene. The Forest Service ships more than a million seedlings annually from the nursery for plantings at former logging sites and wildfire burns across the Northwest.
Like other crops, seedlings grow best when their beds are weed free, and the robot developed by a U of I robotics team from the College of Engineering is a useful tool to accomplish that task.
“My job was to develop the sensor that allows the robot to find the weeds before a mechanical arm comes down and kills them,” Wing said.
The weed detection AI model Wing developed for the wheeled robot had to distinguish weeds from tree seedlings.
“The machine vision needed to be sensitive at ground level, so I built an extensive data set and trained it to get improved results targeting weeds,” he said.
Wing, who grew up in northern Idaho, said he attended college after earning a GED as a home-schooled student. He wanted to pursue a computer science degree to be a software engineer like his father but learned quickly that the structured learning environment of college didn’t suit him. His math phobia also didn’t help in his chosen field.
Math is an integral part of what I do now.Kevin Wing, robotics undergraduate
He dropped out, worked in the restaurant and trucking industries for more than a decade before one day realizing a career founded with a college degree may be his best route forward after all.
One afternoon driving home from his job as a food service manager he saw a billboard advertising a computer science program at North Idaho College. “That billboard lit a fire in me,” he said.
Within a day, Wing enrolled, worked to overcome his fear of math by completing online courses at night, and began to excel as a non-traditional college student preparing for a career in computer science. He earned an associate’s degree as a segway into the U of I robotics program in Coeur d’Alene.
“Math was a big struggle for me,” Wing said.
His journey took him from tutorials in remedial algebra to calculus, linear algebra and statistics.
“Math is an integral part of what I do now,” he said.
Two projects in the U of I college senior’s repertoire include the weeding robot project, and another agricultural AI project that uses a field module to monitor weather, temperature, soil moisture and growing conditions for agricultural crops.
“We call it Project Scarecrow,” he said. “Without the 'w' at the end.”
The acronym is SCARECRO, stands for Sensor Collection And Remote Environment Care Reasoning Operation (SCARECRO).
The remote sensing data analysis devices are employed on a scarecrow-like structure enclosed in waterproof electrical boxes erected in agricultural fields. The instruments allow farmers in another state or country to keep an eye on the data – which translates into growing conditions. So far, the AI “scarecrow” has been employed in a winery in Virginia and at U of I’s Sandpoint, Idaho, fruit orchard.
“The most important aspect of the AI winery project is that it collects a massive amount of data,” Wing said. “With the application of various AI algorithms, predictions can be made in regard to the crops, and the data can be accessed from anywhere.”
After he graduates, Wing plans to attend graduate school – U of I offers master’s and doctoral computer science degrees at the Coeur d’Alene campus – and he may consider teaching.
“I love school, and I love learning, and I’d love to pass that on,” Wing said. “So, teaching may be in my future plans and goals.”
