Xiaomeng Xu

My research explores how machine learning can bring unconventional robots to life—robots that don’t just look different, but behave in ways standard robots never could. I develop algorithms that help robots learn to move, sense, and adapt by completely rethinking both their form and control. This includes sensor-rich robots that sense and interact with their entire body, and sensor-less systems that embed intelligence directly into their physical shape. By integrating machine learning with mechanical design and biomechanics, I’m expanding what robots can be and how they learn to interact with the world. This leads to more flexible and intelligent robot systems capable of solving real-world problems that today’s conventional robots cannot handle.