Publications

Planetary ground robots can overturn on crater slopes, uneven terrain, and obstacles. Argus, a family of spherical robots with radially distributed actuators, moves in any body orientation. Across 1,536 simulated morphologies, higher dynamic isotropy was associated with lower tracking error, higher task success, and lower cost of transport, including zero-shot tracking on simulated Mars and Moon terrain under reduced gravity. The physical 20-leg Argus climbed between parallel walls under counterweight-emulated lunar gravity and kept moving with disabled actuators.

We present MARBLE, a fully enclosed omnidirectional amphibious rolling robot driven entirely by internal mass redistribution. Three orthogonal linear sliders shift internal masses to roll a sealed spherical shell that serves as ground contact, buoyant hull, and mount for passive fins, so the same mechanism rolls on land and propels itself on water without reconfiguration, and tolerates direct contact with terrain and obstacles.

We introduce the Cartesian Hand, a 7-DoF end-effector that combines independent grasping and relative manipulation using only linear motion. Two parallel grippers hold different parts of an object while four translating fingertips move them relative to each other, enabling cap opening, pipetting, pumping, screwdriving, trigger actuation, and in-grasp reorientation across 35 objects, on both a robot arm and a humanoid.

We introduce the visible-reachable workspace (VRW), a design-stage measure that conditions visibility on feasible reaching configurations, and apply it to build Duke Humanoid V2, a 31-DoF humanoid with two independently actuated RGB-D cameras. Camera articulation raises visible-reachable coverage from 38% to 97%, and the dual-actuated design cuts two-target reach-and-grasp completion time by 17% and energy by 19% versus the same robot with fixed cameras.

We propose CEER, a compliant end-effector-root (EE-root) control abstraction for modular humanoid loco-manipulation within a hierarchical planning framework. CEER enables compliance-aware whole-body control in an interpretable task space defined by root motion commands and end-effector pose targets, and supports plug-and-play integration with heterogeneous high-level planners. Experiments in simulation and on hardware demonstrate 3.3 cm end-effector tracking accuracy with substantially reduced jerk, stable contact-rich manipulation under teleoperation, and up to 70% success in simulated single-object loco-manipulation tasks within a room-scale environment.

We introduce dynamic symmetry, the uniformity of a robot's attainable center-of-mass accelerations, and show that maximizing it consistently improves trajectory tracking, robustness, and energy efficiency across 1,000+ simulated morphologies. The 20-leg spherical robot we built, Argus, has near-extreme dynamic isotropy. It moves in any orientation, crosses cluttered and deformable terrain, and keeps working when some actuators fail.

We present the Duke Humanoid, an open-source 10-DOF child-sized bipedal robot designed for energy-efficient locomotion using passive dynamics. We develop a reinforcement learning policy deployable zero-shot on hardware for velocity-tracking walking, and propose an end-to-end RL algorithm that encourages passive dynamics, reducing cost of transport by up to 50% in simulation and 31% in real-world tests.

Text2Robot converts user text descriptions into physical quadrupedal robots. Text-to-3D models give diverse starting body shapes within minutes. Within a day, geometric processing and co-optimization of body and control produce a walking robot that accounts for real electronics and manufacturability.

We present Flexipod, an open-source untethered quadrupedal soft robot platform for dynamic locomotion. The robot is 80 vol.% soft with 3D-printed gyroid-infill flexible legs that passively stabilize on multi-terrain environments. With gaits tuned in a CUDA-accelerated soft-body simulator, the real robot achieves 0.9 m/s (2.5 body lengths/sec), faster than most untethered legged soft robots, and can execute backflips.

Adding 8 wt.% diamond nanoparticle filler increases the thermal conductivity of silicone/ethanol soft actuators from 0.190 to 0.212 W/mK, improving actuation speed and cycle count without affecting mechanical properties. An LSTM network predicts the actuation force in multi-cyclic experiments, which makes it possible to model long-term actuator behavior.

We present a multimaterial printer and associated processes for in situ fabrication of silicones and silicone-based elastomer composites for soft actuators. Rheological study reveals conditions for continuous 3D printing of silicone rubber and silicone/ethanol composite. Because the printed material is anisotropic, print modality can be used to tune actuator properties. We print multimaterial soft actuators in a single job.

We developed a virtual reality version of the floor maze test (VR-FMT) to assess spatial navigation in 55 healthy adults across real and virtual environments, with and without maze walls. The VR-FMT gives an ecologically valid assessment of spatial navigation that correlates with executive function, and it may help with early detection of cognitive and navigational deficits.

We examine electrically conductive fabrics as Joule heaters for silicone/ethanol heat-activated soft actuators. A stretchable kirigami-based Thermionyx fabric design withstands up to 195 °C and 30 W, enabling cyclic heating to 80 °C and linear actuation strain exceeding 30%. It outperforms traditional Ni-Cr wire coils, and the approach may carry over to other thermally actuated systems.

Twelve healthy adults walked overground in a virtual environment with continuous multidirectional visual-field perturbations via a VR headset. Participants took shorter strides with higher variability, and medio-lateral perturbations had larger effects than antero-posterior ones. Over time, participants adapted by increasing stride length and reducing variability. This is the first evidence of visuomotor adaptation during overground walking with a VR headset.