Elastra: A Simulation Framework for Elastic Bodies
Organizations: University of Maryland, College Park · The Chinese University of Hong Kong
Abstract
When robots interact with elastic surfaces such as mattresses and trampolines, changes in loading cause the surfaces to deform and rebound, altering subsequent contact locations and forces. We present Elastra, a simulation framework for modeling this interaction and supporting the training and evaluation of robot control policies. Elastra represents each surface using elements that move only vertically, with elastic connections and damping to model deformation and recovery. Contact forces drive surface motion, and the updated surface geometry is used to compute subsequent contact forces. We systematically validate our simulations against an independent implementation and a real mattress load curve. We further use Elastra to train the Unitree G1 humanoid to stand up on mattresses and trampolines. Using the pretrained HoST policy and ProtoMotions motion tracker as original controllers, we train residual policies through reinforcement learning to add corrections to their joint targets. Adding these residual policies increases the success rate with HoST from 44.1% to 72.5%, and from 55% to 89% on the trampoline. With ProtoMotions, the success rate increases from 19.0% to 45.0% for initial poses not encountered during training. The code and video are available on https://github.com/yichg/elastra.
Figures & tables
| Mattress | Trampoline | |
|---|---|---|
| Grid and size | blocks of m covering m | nodes at 0.1 m spacing, circular region of radius 1.2 m |
| Vertical travel | 0.247 m, rigid floor 0.26 m below the reference top | 0.43 m |
| Elastic parameters | N/m, N/m, 306 bending terms, both | N/m, N/m 3 |
| Damping | N s/m, damping ratio 0.25, scales with | scales with |
| Inertial mass | 0.02915 kg per block, about 11.1 kg in all | node mass fixed |
| Contact geometry | rigid blocks moving with the cells | flex mesh moving with the nodes, collision radius 2 mm |
| Input | HoST | Tracker | Representation |
|---|---|---|---|
| Joint positions | 23 | 29 | joint angles |
| Joint velocities | 23 | 29 | factor 0.1 |
| Base orientation | 4 | 4 | quaternion |
| Base angular velocity | 3 | 3 | base frame, factor 0.25 |
| Original controller output | 23 | 29 | HoST increment, tracker target |
| Torso height, torso uprightness, pelvis height | 3 | 3 | three scalars |
| HoST | 23 actuated joints, action scale , observation history of six control steps, trial 5.02 s, correction bound 0.33 rad (1.32 action units) |
| ProtoMotions | 29 actuated joints, clips of 425, 361 and 497 control steps (8.50, 7.22, 9.94 s), heading aligned at the start, correction bound 0.85 rad |
| Time steps | simulation step 0.625 ms, control period 20 ms (32 substeps) |
| Standing condition | pelvis height above the support below the lowest robot point m, base uprightness , held 50 control steps (1 s) |
| Reward weights | , , |
| Networks | policy and value, two hidden layers of 256 and 128, tanh, orthogonal init gain , policy output gain 0.01, initial log std |
| PPO | 256 environments 64 steps per update, 4 passes 4 minibatches, 1000 updates for HoST in two stages of 500 and 500 updates for ProtoMotions (16,384,000 and 8,192,000 steps), , GAE 0.95, clip 0.2, lr , value coef 0.5, entropy 0, grad norm 1.0, target KL 0.05 |
Robotics Simulation
Robotics simulation builds virtual robots and environments to model, train, and evaluate autonomous systems before real-world deployment.
See all 349 papers