MocoBot operates during nighttime hours when pests are most active. It captures specialized low-light images and uses AI trained to recognize pests in various poses and conditions. This allows the robot to function effectively in real-world farming environments.
"We realized a lot of farmers can’t afford high-tech robots," Choi said. "We’re committed to keeping MocoBot cost-effective by using inexpensive robotic platforms and simplistic AI models. This, in turn, ensures that performance remains high without requiring costly, high-powered hardware.”
The development process included a three-phase training regimen: teaching the robot to identify pests, programming a robotic arm to remove them, and enabling safe navigation through fields without harming crops. Choi’s team collaborated with local farmers and the KSU Field Station to build a diverse image dataset, as no existing nighttime pest database was available.
Choi sees MocoBot as the beginning of a broader initiative to apply AI and robotics to sustainable agriculture. “Food security is a growing concern as the global population increases,” Choi said. “By helping farmers reduce pest-related losses without harming the environment, we can make a real difference.”
CCSE Dean Sumanth Yeduri praised the project, saying, “Dr. Choi's research on MocoBot is a perfect example of how our faculty are pushing the boundaries of technology to solve real-world problems.”