Hardware-as-Regularization: Enhancing Data Consistency in Imitation Learning via an Adjustable Passive Compliant Leader Arm
Low-cost passive leader arms enable scalable data collection for imitation learning, but their lack of restoring forces leaves redundant joints unconstrained, yielding highly inconsistent joint-space trajectories.
ProofPaper ↗
Key points
- We address this with a low-cost passive joint module (an adjustable mechanical clutch and a 3D-printed spiral spring) that pulls each joint toward a user-defined neutral posture.
- This passive restoring force acts as a physical prior equivalent to joint-space $L2$ regularization ("Hardware-as-Regularization"), improving kinematic consistency with shorter completion time than an unregularized arm and preserved movement smoothness.
- Experiments show substantial variance reduction over a naive passive arm, trajectory consistency on par with an active virtual spring of the same actuator class, and substantially higher downstream success rates (e.g., 73.3% vs. 36.7% on Block Stacking); a stiffness sweep shows consistency degrading monotonically as stiffness decreases, confirming that mechanical stiffness behaves like a regularization weight.
Sources (1)
- [1]Hardware-as-Regularization: Enhancing Data Consistency in Imitation Learning via an Adjustable Passive Compliant Leader ArmarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 16, 05:56 AM
Low-cost passive leader arms enable scalable data collection for imitation learning, but their lack of restoring forces leaves redundant joints unconstrained, yielding highly inconsistent joint-space trajectories.
We address this with a low-cost passive joint module (an adjustable mechanical clutch and a 3D-printed spiral spring) that pulls each joint toward a user-defined neutral posture.
Extractive summary: sentences quoted from the sources.
Before this
- Oct 16, 2026MambaAct: Efficient Mamba Action Experts for Flow-Matching Vision-Language-Action Models
- Oct 15, 2026TIPS: A Tiny Inference-time Policy for Scaling in Diffusion Language Models
- Oct 15, 2026ETH: Data-Balanced Heterogeneous Pipeline Parallelism for Training Large Models
- Oct 15, 2026Fast-VLA: You Only Think Once for Robot Manipulation
- Oct 15, 2026DST-Net: Decomposed Spatio-Temporal Network for Time Series Forecasting
- Oct 15, 2026Adaptive Randomized Neural Networks for PDEs: Structure-Aware Feature Redistribution and Residual-Driven Block Enrichment