In the context of the rapid development of automation and robotics technology, the 2024 Asia-Pacific Robot Creativity Competition, held in Vietnam, has become an exciting platform for students and young engineers to showcase their creativity and technical skills. One of the biggest challenges in the competition is the design and construction of a robot capable of performing a specific task, which is rice harvesting. This task requires not only high precision but also the ability to operate effectively in a diverse and complex competition environment. This paper presents the results of applying Proteus software in designing and constructing control circuits for the actuators of the rice-harvesting robot. The results show that the control system is precise, and the robot's ability to function in a complex and flexible spatial environment meets the competition's requirements, opening up opportunities for the development of autonomous robots that can improve harvesting efficiency in the agriculture sector.

This paper presents a novel framework integrating the Swin Transformer, Polarized Self-Attention (PSA), and SegNext to advance medical image segmentation. The Swin Transformer leverages hierarchical self-attention to effectively model both global and local semantic relationship, addressing the challenges of long-range dependencies in medical imagin-

g. PSA further advance the framework by preserving high resolution in both spatial and channel dimensions, o-

ptimizing pixel level feature for segmentation tasks. SegNext contributes lightweight architectural elements, ensuring computational efficiency without sacrificing accuracy. The proposed approach is comprehensively evaluated on Synapse dataset, achieving state-of-the-art performance in multi-organ tasks. This unified framework sets a new standard for precision and efficiency in medical image segmentation.