Page 105 - Read Online

Zhang et al. Intell Robot 2022;2(3):275­97 I http://dx.doi.org/10.20517/ir.2022.20 Page 293

B, C Mini B, C, A1 A1 A1 A1
ANYmal ANYmal MIT Cheetah ANYmal ANYmal Unitree Unitree Quadruped Humanoid Transformer ANYmal-C ANYmal Unitree Unitree

RaiSim IsaacGym IsaacGym Gym Issac Gym Issac RaiSim MuJoCo Pybullet Pybullet

Velocities, Stability. Deviation, State Motion Pose, Base Limits, Airtime. Motion, Collisions, Velocity Prior Velocity Joint and Foot Mo- Joint Smooth- CoM, End-Effector Quaternions, Track- Dribbling: Torques, Sphere

Base Slip, State Recovery: Foot Smoothness. Joint Rate, Joint Rate, Angular Motion Angular Position. Motion, Collisions, Constraint, Slip. Tracking Walking: Ball Positions. Motor Collected Base Position, Desired Direction, Energy.
Foot State. (Planning), Joint), Tracking, Action Tracking, Action and and and Pose, and Body Torque, Velocities, Body Draw. Ball and Consumed

Planning: Torque, Adaption: Robot Space (Foot, Velocity Self-Collision, Torques, Velocity Torques, Airtime. Feet Linear Tracking, Discrimination Linear Tracking, Velocity Velocities, Clearance, and tion ness, Imitation: Joint Positions, Current Velocity. ing Tracking Distance, Survival, Count. and

Coordi- Adaption: Recov- Joint Joint Joint De- for Joint Latent Low- Joint Desired Joint Joint 1)

Planning: nates. Torques. Joint Desired ery: Positions. Desired Positions. Desired Positions. Torques Joint Positions. sired Desired Positions. Phase Offset, Joint Po- Target. sition High-Level: Command. Level: Positions. Desired Positions. Desired Positions. Table to Package

Joint States, Map. Goal, Feet Recov- Posi- Goal Joint Ter- Actions, Ori- Velocities, Actions. and Velocity, Mo- and Phase, CPG States, Air- Forces, Latent Ve- and Velocity End- Position, ID. Pre- Im- and An- Joint Feet Velocities, and (Supplement Open-Source

State, Elevation State, Map. Position, JointAnglesandVelocities,Grav- ity, PreviousActions, Goal Veloc- Gravity, Previous BaseStateandVelocities, Gravity, and Pose History, Contact Motion, States Goal Ball Clip Joint Rotations, (3-step), depth). (3-dim), (4-dim),

Base Velocity, Base Elevation Velocity. Velocities, Previous Measurements. and Angles and Positions Positions. Base Samples, External and States Joint Gravity, Position, Position, Actions dense Angular Vector Velocity. publications
Planning: Goal Adaption: Torques, Joint and tions Base Motion, Joint entations Joint Wheel Command, Joint tion, Height Friction, time. Base Command, locities, Effector Orientation, vious (4 Orientation and Contact Base

rain
gles
ery:
and
about
age
ity.
TD3, PPO PPO PPO PPO PPO V-MPO, PPO SAC information
SAC, GCPO [28] MO-VMPO,

More Gap
data- terrain. learned prior- for so- using and lever- states loco- con- tra- 2. Reality

and controller framework generation functions motion allow to switchable locomotion exteroceptive perception. locomotion robots human method for approach foot Table to
model-based drivenapproachforquadrupedal uneven over Cheetah agility. record training policy fast reward rewards captures. approach discretely integrating proprioceptive reusable legged of knowledge movement. RL proprioceptive observations control. RL-based evolutionary generator. Solution

unified locomotion Mini MIT achieving robotic achieving parallelism. Substituting stylish motion adversarial RL multiple, quadrupedal Learning real for animal end-to-end both visual motion novel an taining jectory

A A A via with from An based styles. A lution and skills prior An aging and A

Transactions on RSS CoRL ArXiv ArXiv Robotics ArXiv ICLR Autom
IEEE Robotics 2022 2022 2022 2022 2022 Science 2022 2022 2022 IEEE Robotics 2022

Others
Loco- Con- Using Ad- locomo- the in Learn- Movement Animal quadrupedal cross- Trajectory for

Legged Optimal Reinforcement Minutes DRL [35] Multiple RL [56] in perceptive robots Repurpose: and with Approach Locomotion [62]

Terrain-Aware and RL using via Locomotion in Walk to Parallel AdversarialMotionPriorsMakeGood Substitutes for Complex Reward Func- through Skills Priors Motion robust quadrupedal and Robot Reusable Human From vision-guided end-to-end transformers [13] Evolutionary General A

RLOC: motion trol [42] Rapid Learning [36] Learning Massively tions [91] Advanced versarial Learning for tion wild [8] Imitate ing Skills Behaviors [12] Learning locomotion modal with RL Generator: Quadrupedal Publication