Jue's Lab designs intelligent, bio-inspired soft robotic systems — where the body actively morphs, senses, and adapts through smart structures and integrated electronics, and the brain learns, controls, and decides through AI by leveraging the intelligence of the body.
Our lab focuses on building intelligent, bio-inspired soft robotic systems by combining embodied AI, physics-informed AI, mechanical intelligence, advanced manufacturing, and flexible electronics.
More specifically, we develop soft robots with co-intelligence at both the body and brain levels: the body can actively morph, sense, and adapt through smart structures and integrated electronics, while the brain can learn, control, and make decisions through AI by leveraging the intelligence of the body.
The lab is part of the University of Tennessee, Knoxville's Foundational AI Cluster, and looks forward to new collaborations across robotics, AI, manufacturing, and bio-inspired systems.
We pursue soft robots whose physical bodies and AI brains co-evolve — spanning materials, manufacturing, electronics, modeling, and learning.
Reinforcement learning and control that exploit body dynamics for agile, robust locomotion and manipulation.
Sim-to-real and surrogate models grounded in mechanics for data-efficient, generalizable control.
Smart structures that compute, sense, and adapt — offloading intelligence into the body itself.
Array-based actuators and machine learning for universal, controllable 3D shape transformation.
3D printing and direct-ink-writing of multifunctional soft actuators, sensors, and electronics.
Integrated, fabric-embedded sensing and actuation for adaptive, wearable, and bio-integrated systems.
LUART (Learning Untethered Amphibious Robotic Turtle) actively morphs its limbs between inflated, stiff legs for walking and deflated, flat flippers for swimming, so a single robot can move on land, in water, and across the transition between them.
Effective leg length and friction encode soft limbs in a rigid-body simulator, and body-height control compensates for softness, giving agile, adaptive gaits with real-time cost-of-transport (COT) monitoring.
Pressure-regulated waterproof hardware works from the surface to 10 m depth. Gaits are learned with reinforcement learning from hand-designed seed gaits, moving toward full 3D attitude and speed control.
Morph–Energy–Path–Gait (MEPG) planning decides where and when to morph, minimizing the total energy of a route, including the transitions themselves.
Amphibious robotic turtle — learned locomotion (RL + sim2real)
Calibration-free fabric wearable motion capture
Programmable shape-morphing surfaces
Machine-learning–controlled morphing
Universal 3D shape-morphing robotsThe lab is just getting started — and we're growing. Come build it with us.
Fully-funded PhD positions starting 2027. Backgrounds in robotics, ML, ME, EE, or materials are all welcome.
Apply →Postdoc positions are open. If you are interested, please reach out to jw4556@princeton.edu.
Apply →Motivated master's and undergraduate researchers, and visiting scholars, are welcome to reach out.
Get in touch →We are recruiting curious, driven people excited about soft robotics, embodied / physics-informed AI, and bio-inspired systems.
Fully-funded positions starting 2027. Email your CV, a short statement of interest, and (if available) transcripts.
Hands-on research in robotics, manufacturing, electronics, and AI. Reach out with your interests.
We are hiring postdocs. If you are interested, please reach out to jw4556@princeton.edu. Collaborators are also welcome to get in touch.
To apply or inquire, email jw4556@princeton.edu with the subject line "Joining Jue's Lab".