REVIEW 5 cited by
Bridging the Sim-to-Real Gap for Athletic Loco-Manipulation
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
Signed reviews
read the original abstract
Achieving athletic loco-manipulation on robots requires moving beyond traditional tracking rewards - which simply guide the robot along a reference trajectory - to task rewards that drive truly dynamic, goal-oriented behaviors. Commands such as "throw the ball as far as you can" or "lift the weight as quickly as possible" compel the robot to exhibit the agility and power inherent in athletic performance. However, training solely with task rewards introduces two major challenges: these rewards are prone to exploitation (reward hacking), and the exploration process can lack sufficient direction. To address these issues, we propose a two-stage training pipeline. First, we introduce the Unsupervised Actuator Net (UAN), which leverages real-world data to bridge the sim-to-real gap for complex actuation mechanisms without requiring access to torque sensing. UAN mitigates reward hacking by ensuring that the learned behaviors remain robust and transferable. Second, we use a pre-training and fine-tuning strategy that leverages reference trajectories as initial hints to guide exploration. With these innovations, our robot athlete learns to lift, throw, and drag with remarkable fidelity from simulation to reality.
Forward citations
Cited by 5 Pith papers
-
Sampling-Based System Identification with Active Exploration for Legged Robot Sim2Real Learning
SPI-Active identifies legged-robot physical parameters via massive parallel sampling and uses Fisher-information-optimal command sequences to collect informative real-world data, improving sim-to-real transfer on quad...
-
FACET: Force-Adaptive Control via Impedance Reference Tracking for Legged Robots
FACET trains legged robots to track a virtual mass-spring-damper reference, so the user can tune stiffness and virtual mass to control how the robot yields to or applies forces.
-
NeuralActuator: Neural Actuation Modeling for Robot Dynamics and External Force Perception
A Transformer-based actuator model learns torque surrogates and external forces on low-cost servo arms from pose trajectories and motor telemetry, improving force estimation and behavior-cloning control.
-
Thor: Towards Human-Level Whole-Body Reactions for Intense Contact-Rich Environments
A decoupled whole-body RL policy with a force-based lean reward enables a Unitree G1 humanoid to pull with up to 167.7 N, beating prior controllers by 69–75%.
-
Towards bridging the gap: Systematic sim-to-real transfer for diverse legged robots
PACE fits a compact set of actuator parameters from brief in-air data and trains energy-aware locomotion policies that transfer zero-shot to real quadrupeds without dynamics randomization.
Discussion (0). Continue with ORCID to comment.