Why Create RobotFlow?

When we run into the field of embodied AI, or robot learning, we find that the path is not as clear as computer vision. Then, we decide to see what the big-name research labs are working on. But we find they seem to have disputes on almost every subjects of embodied AI.

Thus, it natually occurs to us that we may as well develop and maintain a robot system.

Why Use RobotFlow?

RobotFlow project starts as an inside project for our team to do research. When we collaborate with other teams from different universities and institutes, we find they also show interest in our system. It could save their time of developing the software infrastructures from scratch to involve in the exciting field of embodied AI/robot learning.

RobotFlow provides full-stack solutions for robot learning, including but not limited to multi-modal sensor design and fabrication, perception algorithms for robots, task and motion planning libraries, control suite, and simulation environments.

Aside from the toolkit, we are more interested in develop an open community. When we are fresh learners in this field, we find that many lessons and experiences are not easily accessible in the Internet. Thus, we are willing to fill this gap to create such an open community to share the issues, discuss the ideas for developers and researchers.