RealSense新品介绍路演活动

活动时间

活动介绍

眼见为实：感知突破将如何解锁机器人物理交互的新高度

Seeing to Believe: How Perception Breakthroughs Will Unlock New Heights in Robotic Physical Interaction

机器人技术已经跨越了一个关键的临界点，但真正有意义的物理交互依然受限于一个未受足够重视的瓶颈：感知。一个无法可靠解读周围环境的机器人，是无法让人放心地让它去导航、操作物体，或是与人类安全共事的。从实验室里的技术展示走向真实世界的落地部署，这条路直接取决于其视觉智能的质量、延迟和鲁棒性。

Robotics has crossed a critical threshold, yet meaningful physical interaction remains constrained by a bottleneck that receives too little attention: perception. A robot that cannot reliably interpret its environment cannot be trusted to navigate, manipulate objects, or work safely alongside people. The path from laboratory showcase to real-world deployment runs directly through the quality, latency, and robustness of its visual intelligence.

本次演讲将探讨 3D 视觉、视觉-语言-动作（VLA）模型以及世界模型的进步，是如何为智能机器打造出“视觉皮层”的——赋予机器人在动态环境中感知、适应和采取行动所需的“空间理解力”。中国高度整合的机器人产业生态，为这些技术提供了一个强大的试验场，正加速推动各项突破，从而解锁机器人物理交互的全新高度。

This presentation explores how advances in 3D vision, Vision-Language-Action models, and world models are creating a visual cortex for intelligent machines—giving robots the spatial understanding needed to perceive, adapt, and act in dynamic environments. China’s integrated robotics ecosystem provides a powerful proving ground for these technologies, accelerating the breakthroughs that will unlock new heights in robotic physical interaction.