Robot planning-motion division of labor… response speed gap in Anthropic evaluation
Original headline로봇 계획·동작 분업…앤트로픽 평가서 응답속도 격차
Not investment advice. Muhabbit scores how much a story may matter, not which way to trade.
Dyna Robotics unveiled Dyna-2.1 on September 29, 2026, a split-architecture robot system that separates long-horizon task planning from motion execution, comprising a vision-language model orchestrator, the Dyna-2 policy, and a separate whole-body controller that regulates joint targets and wheel speeds at 100 Hz. The orchestrator logs task state such as whether a washing machine is running and how many towels have been folded, feeding that text memory into subsequent decisions to maintain continuity across long tasks. The company's technical documentation cautions that this memory only preserves progress and does not by itself mean the motion policy has learned new skills. The design matters because decoupling slow planning from fast control could improve responsiveness and reliability in extended manipulation tasks, a key bottleneck for commercial robotics. What to watch next is independent benchmarking—including the cited Anthropic evaluation, where response-speed gaps reportedly appeared—and whether the modular approach translates into measurable gains in real-world deployment.
Muhabbit's AI wrote this summary in its own words from TokenPost's report and translated it from Korean. It can get details wrong, so check the source.
Named in this story
Dyna Robotics, Dyna-2.1, Anthropic