Development of software for inference board

This project introduces an inference board solution designed for mobile robots. As robotic systems grow in complexity, the demand for efficient, cost-effective computing power increases. Our inference board addresses this challenge by offloading intensive AI tasks from the main robot computer. This approach not only enhances overall system performance but also allows for greater scalability and energy efficiency. By utilizing low-cost hardware and optimized software, the solution maintains high performance while significantly reducing production costs. The board seamlessly integrates with existing robot architectures, supporting both ROS1 and ROS2, making it versatile for a wide range of robotic applications. Whether for advanced perception tasks or complex decision-making processes, this inference board provides the necessary computational power without compromising speed or accuracy.

Jiwon Park
Jiwon Park

My research interests include perception for robust off-road mobile robot