$\chi_{0}$: Resource-Aware Robust Manipulation via Taming Distributional Inconsistencies

2026-02-09

AI Summary

This work addresses error accumulation in long-horizon robotic manipulation caused by distributional mismatches among human demonstrations, policy training, and test-time execution. To this end, the authors propose the χ₀ framework, which fuses multi-source demonstration distributions via model arithmetic, introduces a phase-aware advantage estimator to deliver stable reward signals, and aligns training and deployment distributions through spatiotemporal augmentation, heuristic DAgger, and temporal chunk smoothing. This approach is the first to systematically identify and mitigate this triple distribution shift. Evaluated under resource constraints—using only 20 hours of demonstration data and eight A100 GPUs—the method achieves a task success rate 3.5× higher than the state-of-the-art π₀.₅ baseline, enabling robust 24-hour autonomous operation of a dual-arm robot for complex tasks such as garment unfolding, folding, and hanging.

Technology Category

• Intelligent Robots: Manipulation
• Multiagent Systems: Adversarial Agents
• Planning, Routing, and Scheduling: Learning for Planning and Scheduling

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

High-reliability long-horizon robotic manipulation has traditionally relied on large-scale data and compute to understand complex real-world dynamics. However, we identify that the primary bottleneck to real-world robustness is not resource scale alone, but the distributional shift among the human demonstration distribution, the inductive bias learned by the policy, and the test-time execution distribution -- a systematic inconsistency that causes compounding errors in multi-stage tasks. To mitigate these inconsistencies, we propose $\chi_{0}$, a resource-efficient framework with effective modules designated to achieve production-level robustness in robotic manipulation. Our approach builds off three technical pillars: (i) Model Arithmetic, a weight-space merging strategy that efficiently soaks up diverse distributions of different demonstrations, varying from object appearance to state variations; (ii) Stage Advantage, a stage-aware advantage estimator that provides stable, dense progress signals, overcoming the numerical instability of prior non-stage approaches; and (iii) Train-Deploy Alignment, which bridges the distribution gap via spatio-temporal augmentation, heuristic DAgger corrections, and temporal chunk-wise smoothing. $\chi_{0}$ enables two sets of dual-arm robots to collaboratively orchestrate long-horizon garment manipulation, spanning tasks from flattening, folding, to hanging different clothes. Our method exhibits high-reliability autonomy; we are able to run the system from arbitrary initial state for consecutive 24 hours non-stop. Experiments validate that $\chi_{0}$ surpasses the state-of-the-art $\pi_{0.5}$ in success rate by nearly 250%, with only 20-hour data and 8 A100 GPUs. Code, data and models will be released to facilitate the community.

Problem

• distributional shift
• robust manipulation
• long-horizon robotic tasks
• compounding errors
• real-world robustness

Innovation

• Distributional Consistency
• Model Arithmetic
• Stage Advantage
• Train-Deploy Alignment
• Resource-Efficient Robotic Manipulation