Exoskeletons and wearable robots
Exoskeletons and exosuits that assist walking and lifting, and how their assistance is tuned to each person.
Start here: seminal papers
Reducing the energy cost of human walking using an unpowered exoskeleton★ Must read
Nature, June 2015
An ankle exoskeleton with no motor or battery, just a spring that a mechanical clutch engages while the foot is on the ground, lowered the metabolic cost of walking. With the best of five springs, net metabolic rate across nine participants fell 7.2 ± 2.6% compared with walking without the device.
Feedback Control of an Exoskeleton for Paraplegics: Toward Robustly Stable Hands-free Dynamic WalkingSeminal
February 2018
The authors port hybrid-zero-dynamics control from bipedal robots to the 12-actuator ATALANTE exoskeleton so complete paraplegics walk without crutches. Three patients walked about 9-10.5 m hands-free at 0.11-0.15 m/s, and a newer learned controller recovers from 750 N pushes in simulation where a fixed gait fails above 300 N.
Towards Variable Assistance for Lower Body ExoskeletonsSeminal
September 2019
A control-barrier-function filter lets an exoskeleton wearer move the swing-leg hip and knee freely inside a tube around the nominal gait and intervenes only at the tube boundary. On the Atalante exoskeleton with eight able-bodied subjects, lowering the assistance (a wider tube) raised the metabolic power the wearer had to contribute.
Preference-Based Learning for Exoskeleton Gait OptimizationSeminal
September 2019
CoSpar picks an exoskeleton's gait from a library of precomputed stable gaits using only pairwise preferences ("better than the last trial?") and suggestions like "longer step". In 20 trials per user on the Atalante exoskeleton, its learned preferences matched the blind rankings of all three subjects.
The exoskeleton expansion: improving walking and running economySeminal
Journal of NeuroEngineering and Rehabilitation, February 2020
A review of the 23 studies, through December 2019, in which a lower-limb exoskeleton made walking or running cheaper than going without one. On level ground the devices cut net metabolic cost by 3.3-19.8% for walking and 3.9-8.0% for running.
Human Preference-Based Learning for High-dimensional Optimization of Exoskeleton Walking GaitsSeminal
March 2020
LineCoSpar learns a person's preferred exoskeleton walking gait from pairwise comparisons by searching one random line of the gait parameter space at a time. On the Atalante exoskeleton it tuned six gait parameters for six able-bodied users in 30 trials each.
Personalizing exoskeleton assistance while walking in the real world★ Must read
Nature, October 2022
A model that ranks ankle-exoskeleton torque patterns from the wearer's ankle motion replaces lab respirometry, so a portable exoskeleton can personalize its assistance during everyday walking. After one hour of optimization on a public sidewalk, assistance cut the energy cost of transport by 17% and raised walking speed by 9% compared with normal shoes.
Recent
ExoStack: A Modular and Reconfigurable Software Framework for Wearable Robotics Using ROS 2
November 2026
An open ROS 2 software stack that runs the whole wearable-robot loop, from ESP32 sensor hubs to gait-event detection, assistance profiles and human-in-the-loop optimization, configured through YAML files instead of new code. The same code runs two different pneumatic exosuits, and eight hubs stream at 99.5 Hz against a 100 Hz target.
Compact and Wide-Force-Range Magnetorheological Hand Exoskeleton for Kinesthetic Feedback in Teleoperation
November 2026
A hand exoskeleton with a 39.8 g magnetorheological brake per finger gives kinesthetic feedback while teleoperating a robot hand. It renders fingertip forces up to 21 N with about 1 N of back-driving force, and 12 users gripped more gently with it (3.50 N vs 4.61 N).
Enhancing Balance and Posture in Material Handling Tasks with a Wearable Robotic Tail
November 2026
A wearable robotic tail swings a 3 kg counterweight behind the user to offset a load held out in front. In a 10-person shelf-loading study it cut peak centre-of-pressure displacement by 17-20 mm and peak trunk inclination by 4.3 degrees compared with no tail.
Multi-Objective Human-in-the-Loop Bayesian Optimization of a Lower-Limb Exoskeleton
September 2026
MO-HILBO uses multi-objective Bayesian optimization to learn each wearer's set of Pareto-optimal hip exoskeleton controllers, trading metabolic cost against self-rated comfort. With 24 three-minute trials per person (3 subjects), its Pareto controllers dominated anti-Pareto ones in validation, and it predicted 17 of 18 pairwise orderings of held-out controllers.
Learning a Speed-adaptive Hip Exoskeleton Control Policy Via Sim-to-real Reinforcement Learning
September 2026
They learn when a hip exoskeleton should push with sim-to-real RL on a musculoskeletal model walking at varying speeds, then tune only two torque gains on the real device from users' pairwise preferences. In free walking it cut negative mechanical power by 55% compared with a delayed-feedback controller.
UniExo: Unified Multi-Skill Policies for Musculoskeletal Locomotion and Co-Adaptive Exoskeleton Control
September 2026
UniExo distills four musculoskeletal locomotion experts into one skill-latent human policy that tracks 94.7% of unseen clips, then co-trains a single hip exoskeleton controller with it in simulation. On hardware the controller assisted six people at four treadmill speeds and one person through walking, turning, running and backward walking without mode switching.
Musculoskeletal Motion Imitation for Learning Personalized Exoskeleton Control Policy in Impaired Gait
April 2026
An RL policy drives a 90-muscle musculoskeletal model to imitate human walking and running, then is fine-tuned to also command hip or ankle exoskeleton torques, including for simulated muscle weakness. In simulation, ankle assistance cuts metabolic cost by 13.8% in running, and deficit-specific assistance cuts it by 8.1-13.1% in impaired gait.
Learning Hip Exoskeleton Control Policy via Predictive Neuromusculoskeletal Simulation
March 2026
Train a hip-exoskeleton assistance policy entirely in a muscle-driven human walking simulation, with no motion capture, then distill it into a thigh-IMU controller on real hardware. In simulation the assistance cuts muscle activation by up to 3.4% and positive joint power by up to 7.0%, and on hardware the torque profiles match simulation with r = 0.82.
Exo-Plore: Exploring Exoskeleton Control Space through Human-aligned Simulation
January 2026
Exo-plore tunes a hip exoskeleton's gain and delay entirely in a muscle-driven simulation of a walker trained with deep RL, with no human walking trials. Its assisted gaits follow human trends (0.96 correlation on RMS assistive moment), and the optimal gain varies linearly with severity in four of five simulated gait pathologies.
Improving CMA-ES Convergence Speed, Efficiency, and Reliability in Noisy Robot Optimization Problems
January 2026
AS-CMA adds per-candidate sample-time selection to CMA-ES, measuring longer only when a candidate is predicted to be hard to rank against its neighbors. On four simulated noisy landscapes it reached fine convergence in 98% of runs, 24-65% faster than the best fixed sample time.
New papers stay for 14 days.
Summaries are written automatically from each paper's text. Figures and videos come from the paper and its project page; all credit goes to the authors.