SplitAdapter: Load-Aware Humanoid Loco-Manipulation via Factorized Adaptation
2026-06-02 • Robotics
Robotics
AI summaryⓘ
The authors address the challenge of controlling humanoid robots that need to pick up and place objects of different weights and at different heights, which is especially tricky when moving from simulation to the real world. They introduce SplitAdapter, a method that improves the robot’s control by separately understanding the object’s weight and the robot’s own dynamics, rather than combining them into one. Their approach uses special training techniques to make these components work together better, resulting in more reliable manipulation, particularly with heavier objects. Tests in both simulated and real environments showed that SplitAdapter helps the robot succeed more often than previous methods.
Humanoid robotLoco-manipulationSim-to-real transferLoad variationContext encoderFeature-wise Linear Modulation (FiLM)Cross-adversarial regularizationWorld-modelGrip manipulationHierarchical adaptation
Authors
Jeonguk Kang, Hanbyel Cho, Sanghyun Kang, Donghan Koo
Abstract
Humanoid loco-manipulation requires stable whole-body control under varying object masses and pickup/placement heights. This becomes particularly challenging in sim-to-real transfer, where object-induced load variation and robot-side dynamics mismatch interact during physical contact. Existing history-based adapters often compress these factors into a single latent representation, which can weaken robustness under heavy-load manipulation. We propose \textbf{SplitAdapter: Load-Aware Humanoid Loco-Manipulation via Factorized Adaptation}, which freezes a pretrained box manipulation policy and extends it with object/load and dynamics-aware context encoders trained with split world-model objectives, GRL-based cross-adversarial regularization, and hierarchical Feature-wise Linear Modulation (FiLM). In sim-to-sim experiments and real-world deployment, SplitAdapter improves Full-task success over the base policy and world-model FiLM baselines across object masses of $2$, $4$, and $6$ kg and pickup/placement heights of $0$, $30$, and $60$ cm, with the largest improvements under heavy-load conditions.