Robot hardware and actuators
New robot designs, from hands and arms to legged and humanoid platforms, and the actuators and transmissions that drive them.
Start here: seminal papers
Quasi-Direct Drive for Low-Cost Compliant Robotic Manipulation★ Must read
April 2019
Blue is a human-scale 7-DoF arm built from quasi-direct-drive motors and timing belts, so it is backdrivable and force-controlled without torque sensors. It carries 2 kg, has a 7.5 Hz nominal position bandwidth and 4 mm repeatability, and is projected to cost under $5000 at volumes above 1500 arms.
Stanford Doggo: An Open-Source, Quasi-Direct-Drive QuadrupedSeminal
May 2019
Stanford Doggo is a 4.8 kg open-source quadruped driven by 3:1 belt-drive quasi-direct-drive actuators that costs under $3000 to build. It jumps 1.14 m and reaches a vertical jumping agility of 2.23 m/s, 22% above the previous best robot.
An Open Torque-Controlled Modular Robot Architecture for Legged Locomotion Research★ Must read
September 2019
An open-source, torque-controlled quadruped (Solo) built from eight identical 3D-printed actuator modules with brushless motors and a 9:1 belt transmission. The 2.2 kg robot cost about 4000 € in materials, reaches a dimensionless leg stiffness of 10.8 and jumps to a base height of 65 cm.
The MIT Humanoid Robot: Design, Motion Planning, and Control For Acrobatic BehaviorsSeminal
April 2021
A design for a 0.7 m, 21 kg humanoid built on MIT Cheetah-style proprioceptive actuators, with a kino-dynamic planner that respects actuator limits and an MPC plus whole-body landing controller. In a simulation that models the measured actuators and battery, it does back flips, front flips and spinning jumps.
Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware★ Must read
April 2023
ALOHA is a bimanual teleoperation rig built from off-the-shelf arms for under $20k, and ACT is an imitation learning policy that predicts chunks of future joint targets with a transformer CVAE. From 50 demonstrations per task, ACT slots a battery 96% of the time and opens a translucent condiment cup 84% of the time.
Getting the Ball Rolling: Learning a Dexterous Policy for a Biomimetic Tendon-Driven Hand with Rolling Contact JointsSeminal
August 2023
Introduces the Faive Hand, a 3D-printed tendon-driven hand with rolling contact joints, and models those joints in IsaacGym to train an RL policy. After about one hour of training on a single GPU, the policy transfers zero-shot to the real hand and rotates a sphere in-hand using only proprioception.
LEAP Hand: Low-Cost, Efficient, and Anthropomorphic Hand for Robot LearningSeminal
September 2023
A four-finger anthropomorphic robot hand built from Dynamixel motors, off-the-shelf brackets and 3D-printed parts in about 4 hours for 2000 USD, 1/8th the cost of Allegro. A new knuckle joint keeps sideways finger motion in every pose, and the hand matches or beats Allegro on 9 of 10 teleoperated tasks.
Mobile ALOHA: Learning Bimanual Mobile Manipulation with Low-Cost Whole-Body TeleoperationSeminal
January 2024
A $32k bimanual mobile manipulator whose operator, tethered to the base at the waist, drives the wheels while puppeteering both arms, plus behavior cloning co-trained on existing static ALOHA data. With 20-50 demos per task, ACT reaches 80-95% success on six of seven household tasks.
Berkeley Humanoid: A Research Platform for Learning-based Control★ Must read
July 2024
A 16 kg, 0.85 m humanoid built in-house for about USD 10,000 and designed so that simulation matches the hardware. A plain MLP policy trained with light domain randomization transfers zero-shot, walking 364 m across campus and hopping on one leg.
TidyBot++: An Open-Source Holonomic Mobile Manipulator for Robot LearningSeminal
December 2024
An open-source holonomic mobile base built from powered casters for $5-6k, carrying a Kinova Gen3 arm and teleoperated with a phone. Diffusion policies trained on 50-100 demonstrations reach 6/10 to 10/10 success on six household tasks, and the holonomic base beats a differential-drive mode 9/10 to 4/10 on countertop wiping.
Design and Control of a Bipedal Robotic CharacterSeminal
January 2025
A 15.4 kg, 0.66 m bipedal robot character whose shape and motion come from animators, controlled by separate RL policies for standing, walking and short animated clips that an operator blends live from a two-joystick controller. Public deployments with up to three robots ran on the order of 10 hours without a fall.
ToddlerBot: Open-Source ML-Compatible Humanoid Platform for Loco-ManipulationSeminal
February 2025
An open-source, fully 3D-printed miniature humanoid (0.56 m, 3.4 kg, 30 active DoFs) that costs under 6,000 USD and is built for collecting both simulation and real-world data. Diffusion policies trained on 60 teleoperated demonstrations reach 90% success on bimanual and 75% on full-body manipulation.
ORCA: An Open-Source, Reliable, Cost-Effective, Anthropomorphic Robotic Hand for Uninterrupted Dexterous Task LearningSeminal
April 2025
An open-source, 17-DoF tendon-driven anthropomorphic hand with fingertip touch sensors that one person can build in under eight hours for less than 2,000 CHF. It ran an imitation-learning pick-and-place policy for over 7 hours (about 2,000 grasps) with no intervention on the hand hardware.
RUKA: Rethinking the Design of Humanoid Hands with LearningSeminal
April 2025
RUKA is an open-source, human-sized, tendon-driven hand with 15 degrees of freedom and 11 actuators that costs under $1,300 to build from 3D-printed and off-the-shelf parts. Its controllers are learned from data recorded by a MANUS motion-capture glove fitted onto the robot hand, and it beats LEAP and Allegro on pinch force, payload (6.0 kg) and slip tests.
Demonstrating Berkeley Humanoid Lite: An Open-source, Accessible, and Customizable 3D-printed Humanoid RobotSeminal
April 2025
An open-source, 0.8 m, 16 kg humanoid built from desktop-3D-printed parts and off-the-shelf components for under $5,000. Its 3D-printed cycloidal actuators survive a 60-hour durability test, and the robot walks with an RL policy transferred zero-shot from simulation.
Recent
Helping Hand: A Continuum Robotic Grasper for Traction in Single-Channel Endoscopic Submucosal Dissection
November 2026
The Helping Hand is a tendon-driven, notched Nitinol continuum grasper with a slot at its base for an electrosurgical knife, so tissue traction and cutting both fit through one 3.7 mm colonoscope channel. It applied about 0.33 N of traction at every tested phantom height, and with the knife it removed a tissue sample from ex vivo chicken thigh in one piece in 15 min.
Actuation and Control of Passive Tendon-Driven Soft Extensions Through Rigid Manipulator Motion
November 2026
An industrial arm bends a passive tendon-driven soft extension with its own joint motion, because tendons anchored on upstream links tighten or slacken as the arm moves, while the end effector holds a position or fulcrum constraint. In simulation the tip reached every feasible target within 4 mm; on a real UR10 with handmade extensions costing under US$10, it reached 6 of 16 targets and every safety barrier held.
RAIPAL: A Powerful and Agile Humanoid Arm for Human Tasks
November 2026
A 7-DoF, 8.9 kg humanoid arm sized from published human upper-limb measurements, with a strong upper arm, a fast forearm and an elbow linkage whose gearing changes with angle. One arm holds a 20 kg payload, pitches at 20.6 m/s endpoint speed and survives 70 hammer strikes.
MABEL: An Open Anthropomorphic Mobile Bimanual Platform for Teleoperation and Learning
November 2026
An open-source 56-DOF mobile bimanual robot with dexterous hands, a pitching torso, a lift and an active neck, for a US$8.1k parts cost. Apple Vision Pro head and hand tracking drives the whole body with 137 ms video latency, and an ACT policy trained through the same controller succeeds in 8 of 10 grasp-and-place trials.
Pixel: A Co-Designed Full-Stack Robot System for Learning General-Purpose Mobile Manipulation
November 2026
Pixel is a wheeled humanoid that costs under $15,000. The robot, its teleoperation interfaces, a wearable human-demonstration rig, the whole-body controller and the policy were designed together, so human and robot data share one action space. Policies trained only on wearable-rig demonstrations reach 82.0% average task progress on three household tasks, and collecting those demonstrations is 4.2-6.2x faster than teleoperation.
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.