Qwen Launches Unified Embodied Control Model
CyberRobooo · x · 2026-07-19
Alibaba's Tongyi Qwen has introduced **Qwen-VLA**, a unified Vision-Language-Action model designed to directly control robots of various forms, including humanoid and dual-arm setups. It integrates manipulation, navigation, trajectory prediction, and cross-form control into a single system, using "embodied perception prompts" to adapt to different robot embodiments without needing to train a separate control head for each form. The post also claims that Qwen-VLA matches or surpasses specialized models on key benchmarks, achieving a **76.9%** Out-of-Distribution (OOD) success rate on real-world ALOHA bimanual tasks.
More from Embodied
- Generative Bionics turns its Gene.01 humanoid from CES concept into a walking platform — CyberRobooo · 2026-07-21
- Auki pitches physical AI infrastructure for stores, warehouses and supply chains — broodsugar · 2026-07-21
- Safe RL paper on directional constraints accepted to IROS 2026 — Jan_R_Peters · 2026-07-21
- Sunday Robotics says one example can teach its home robot a new behavior — saranormous · 2026-07-21
- San Francisco event seeks sponsors for the first 6-foot humanoid robot fight — cixliv · 2026-07-21
- Tiny memristor chip brings brain-model inference under 10 ms — striketheviol · 2026-07-21