BridgeVLA++: A Memory-Augmented 3D Vision-Language-Action Framework for Robotics

ArticlePeiyan Li, Yuze Zhu, Yixiang Chen, Qisen Ma, Yuan Xu, Jiabing Yang, He Guan, Yan Huang, Hongtao Wu, Xiao Ma, Tao Kong, Liang Wang, Tieniu TanResearch

Researchers introduce BridgeVLA++, an upgraded vision-language-action framework designed for 3D robotic manipulation. It introduces a unified spatio-temporal memory architecture into pre-trained vision-language models, enabling persistent spatial reasoning and temporal history tracking without losing data efficiency.

Key points

  1. Existing 3D VLA methods struggle with heavy data requirements, poor generalization under distribution shifts, and a lack of observation memory.
  2. BridgeVLA previously preserved pre-trained VLM alignment by projecting raw point clouds into multi-view images and using intermediate heatmaps.
  3. BridgeVLA++ integrates a spatio-temporal memory architecture to reason across observation history while maintaining generalization.
  4. The framework achieves state-of-the-art results on two memory-dependent benchmarks and transfers successfully to bimanual tasks and real-world robot platforms.

Why it matters

Enabling robots to retain context across sequential actions without requiring massive task-specific training datasets is a major hurdle in deploying versatile, open-world robotic systems.

Analysis

By tackling the absence of explicit memory in standard VLA models, BridgeVLA++ demonstrates how spatio-temporal tracking can be layered onto pre-trained vision-language representations. The reported success across bimanual setups and real robotic hardware highlights its practical scalability for complex physical manipulation.

Mentioned links

Found in

Summary generated by AI (Gemini 3.8 Flash, 27 September 2026). It may contain errors – the original source is authoritative.

Related entries

Similar in content, found by the AI search.

  • Link:Moo Jin Kim, Karl Pertsch, Siddharth Karamcheti et al.

    OpenVLA: Open-Source Vision-Language-Action Model for Robotics

    OpenVLA is an open-source 7B parameter vision-language-action (VLA) model designed for generalist robot manipulation. Built on Prismatic-7B and Llama 2, it was trained on 970k episodes from the Open X-Embodiment dataset and outperforms earlier generalist policies such as Octo, RT-1-X, and in several benchmarks the closed 55B RT-2-X.

    AI & LLMs· Research

  • Article:Moo Jin Kim et al.

    OpenVLA: An Open-Source Vision-Language-Action Model for Robot Control

    Researchers from Stanford, UC Berkeley, Google DeepMind, and partner institutions introduce OpenVLA, an open-source 7-billion-parameter vision-language-action model. Trained on 970k real-world robot demonstrations, it surpasses closed models such as RT-2-X while enabling parameter-efficient fine-tuning on consumer-grade hardware.

    AI & LLMs· Research

  • Link:Zhe Liu, Jinghua Hou, Yuxiang Lu, et al.

    StreamPI: Streaming Multimodal Temporal Modeling for Vision-Language-Action Models

    Researchers from The University of Hong Kong and ACE Robotics introduce StreamPI, a framework that equips single-frame Vision-Language-Action (VLA) models such as π0.5 with temporal reasoning without adding new parameters. By bundling images and instructions into atomic temporal units and using random-interval training, StreamPI improves precision, memory, and robustness in real-robot manipulation.

    AI & LLMs· Research

  • Link:Dyna Robotics

    Dyna Robotics Unveils DYNA-2: World-Action Model Driven by Human Video Data

    Dyna Robotics has launched DYNA-2, a World-Action Model (WAM) pre-trained on more than one million hours of egocentric human video. The model circumvents traditional teleoperation bottlenecks, transferring physical reasoning and action policies directly to diverse robotic hardware.

    AI & LLMs· Announcement

  • Link:arcprize.org

    ARC-AGI-3: Interactive Reasoning Benchmark for AI Agents

    ARC-AGI-3 is an interactive reasoning benchmark designed to evaluate human-like intelligence in AI agents. Moving beyond static puzzles, agents must explore novel environments, derive goals on the fly, and adapt through multi-step interaction.

    AI & LLMs· Resource

Back to the knowledge base

Let's talk about your project

Get in touch

Locations

  • Mattersburg
    Johann Nepomuk Bergerstraße 7/2/14
    7210 Mattersburg, Austria
  • Vienna
    Ungargasse 64-66/3/404
    1030 Wien, Austria

Parts of this content were created with the assistance of AI.