Addressing the Orchestration Gap in Generalist Robots via Physical Agency

A closed-loop VLM orchestrator that drives frozen robot skills — no new robot data, no fine-tuning.

Paper Code orchestrator · thinking

Hold the robot, cameras, scenes, demonstrations, and policy weights fixed. Change only the inference-time process. The same frozen skills succeed far more often inside the loop than when prompted directly — the orchestration gap.

Click a task to watch the agent reason through it — the full video alongside its step-by-step trace.

The orchestrator never emits motor commands. At each step it reads the observation and history, calls one tool, incorporates the result, and decides again — until the task predicate holds. Motion is delegated to skills it never trained.