Hands-On With Astra Robot Control: Johns Hopkins Deep-Dive Tests Its Motion Understanding

jmin__cho · x · 2026-10-09

Lin Long and Jaemin Cho at Johns Hopkins have published a long-form piece systematically evaluating the real-world robot control capabilities of OpenAI's newly released GPT-6 Astra. Astra is hailed as having reached SOTA in robot control performance.

• Highlights: Astra can observe a scene, reason, and directly generate low-level actions, reacting to environmental changes—suggesting a solid grasp of control logic and motion dynamics
• The authors built Ditto-Bench: unlike existing semantics-focused benchmarks, it specifically tests tasks requiring difficult physical interaction
• Core questions: Can Astra anticipate how an instruction will move the robot and affect surrounding objects? Can it account for contact and geometric constraints and adjust when actual motion deviates from expectations?
• The article includes full experiment logs, videos, success and failure cases, with code open-sourced

Original post →

More from Embodied

• Ditto-Bench stress-tests GPT-6 Astra for robot control: simple tasks shine, physics stalls — anand_bhattad · 2026-10-09
• NVIDIA shows frontier AI agents building Omniverse simulations from natural language — nordicinst · 2026-10-09
• NVIDIA Shows Six Projects Building Omniverse Simulations With Frontier AI Agents Like Astra — NVIDIA Blog · 2026-10-09
• Tesla Vision-Only Robotaxi Handles Nighttime Downpours, Silencing Earlier Skeptics — elonmusk · 2026-10-09
• Hark Sees Explosive 48-Hour Growth, Brett Adcock Says Team Scrambling to Fix Infra — adcock_brett · 2026-10-09
• aimotive driving dataset: train labels come from a tracker that sees the future — RexDouglass · 2026-10-09

Embodied channel →