Track: High School STEM Competition
Abstract
This paper explores the application of a robotic arm powered by artificial intelligence (AI) and deep learning to enhance the precision and efficiency of object manipulation tasks. Leveraging OpenAI’s GPT-4o image recognition model along with OpenCV, we implement smart computer vision techniques to accurately recognize, locate, and manipulate targeted objects. By automating control and minimizing human intervention, our approach has the potential for improving traditional kinematics-based methods. The integration of AI enables robots to learn from vision data, adapt to various scenarios, and improve over time. Our findings demonstrate the feasibility of utilizing AI and deep learning to achieve high accuracy and efficiency in robot manipulation, with profound implications for manufacturing automation and beyond.