Scaling VLA Serving Under Robotic Execution Constraints
cs.RO, cs.DC
Submitted: 2026-11-11
Updated: 2026-11-11
Code: https://github.com/X-Square-Robot/xinfer
Terminology
Sources
- Efficient Vision-Language-Action Management and Serving for Robot Factories
- SLOs-Serve: Optimized Serving of Multi-SLO LLMs
- Kairos: A Scalable Serving System for Physical AI
- semi-PD: Towards Efficient LLM Serving via Phase-Wise Disaggregated Computation and Unified Storage
- ROSA: A Robotics Foundation Model Serving System for Robot Factories
- Microscaling Data Formats for Deep Learning
- FlashRT-PAI: Fast Kernels Are Not Fast Pipelines for Physical AI
- Think Twice, Act Once: Token-Aware Compression and Action Reuse for Efficient Inference in Vision-Language-Action Models
- Embodied.cpp: A Portable Inference Runtime of Embodied AI Models on Heterogeneous Robots
- vLLM-Omni: Fully Disaggregated Serving for Any-to-Any Multimodal Models
- Wall-OSS-0.5 Technical Report
- Igniting VLMs toward the Embodied Space
Related papers
- Rethinking Actions: Continuity-Aware Priors for World-Model Planning
- MOSAIC: Modular Obstacle and Spatial-Aware Framework for Manipulation In Cluttered Scenes
- SAGAS: Semantic-Aware Graph-Assisted Stitching for Offline Temporal Logic Planning
- DexSeed: Expanding Human Hand-Object Interactions into Robot Demonstrations for Dexterous Policy Learning
- LaST0.5: A Generalist World Action Model with Interaction-Centric Latent Reasoning
- CIALLO: Learning Shared Intent Representations for Cross-Embodiment Manipulation