arXiv cs.AI· Zixu Hao, Weijun Wang, Xiangyu Li, Liang Mi, Hao Wu, Yunxin Liu, Ting Cao·· 3 小时前AI 评分44
Patchwork：面向域随机化具身仿真的资源共享系统
Replicate What Varies: Resource Sharing for Domain-Randomized Embodied Simulation
AI 导读
研究者提出 Patchwork，一套面向域随机化仿真的资源共享系统，用“共享模板 + 补丁”的表示 E=T+Δ 避免多实例重复复制未变资源。系统在 robosuite 和 MuJoCo 上实现，并在 LIBERO-plus、RoboCasa365 负载上评估：32 个副本下共享模板执行使 CPU RSS 相比每环境一进程方案降低 88-92%。
正文 · 原文
Abstract:Simulation provides a low-cost sandbox where embodied agents can attempt tasks, observe outcomes, and learn from experience. Running many diverse simulation environments concurrently can consume substantial CPU and GPU memory because independently constructed scenes duplicate resources that remain unchanged across instances. We present Patchwork, a resource-sharing system for domain-randomized simulation, built around the representation $E=T+\Delta$: a shared template contains common resources, and a patch records the mutations needed to construct a target environment. Patchwork separates randomization sampling from replay and selects reusable templates together with patches that preserve each target environment's intended configuration. Its layered runtime shares compiled model structure and immutable assets while keeping mutable parameters and simulation state private to each instance. We implement Patchwork on robosuite and MuJoCo and evaluate it on LIBERO-plus and RoboCasa365 workloads. At 32 replicas, shared-template execution reduces CPU RSS by 88-92% relative to process-per-environment execution. Controlled experiments show how sharing shifts memory growth toward template diversity and unique asset content rather than instance count, within the supported sharing boundaries.
| Subjects: | Distributed, Parallel, and Cluster Computing (cs.DC); Artificial Intelligence (cs.AI); Robotics (cs.RO) |
| Cite as: | arXiv:2610.33582 [cs.DC] |
| (or arXiv:2610.33582v1 [cs.DC] for this version) | |
| https://doi.org/10.48550/arXiv.2610.33582 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Zixu Hao [view email]
[v1]
Sat, 24 Oct 2026 10:20:13 UTC (1,925 KB)
来源：arXiv cs.AI · arxiv.org