Search papers, labs, and topics across Lattice.
This paper introduces the Navigation World Action Model (NavWAM), a diffusion-transformer policy that enables robots to execute goal-conditioned visual navigation by integrating future observations, goal progress, and action chunks into a unified latent sequence. By jointly learning future predictions alongside action and value targets, NavWAM effectively transforms visual foresight into actionable control without relying on external planners. Evaluations show that NavWAM outperforms traditional planning-based world models in both offline benchmarks and real-robot deployments, demonstrating its efficacy in real-world navigation tasks.
NavWAM turns visual foresight into executable robot actions, outperforming traditional planning methods in real-world navigation scenarios.
Goal-conditioned visual navigation requires a robot to act under partial observability by anticipating how its motion will change the future egocentric view and whether that change brings it closer to the goal. Navigation world models provide such visual foresight, but they remain prediction modules that require an external planner to convert predicted futures into closed-loop control. We propose Navigation World Action Model (NavWAM), a diffusion-transformer policy that turns navigation world-model prediction into executable action by representing future observations, goal-progress values, and action chunks in a shared latent sequence. By learning future prediction jointly with the action and value targets that determine closed-loop behavior, NavWAM makes visual foresight directly usable for robot control. We build NavWAM through simulation pretraining and real-robot adaptation, and evaluate it on image-goal navigation against planning-based world models and a representative direct navigation policy. Across offline benchmarks and closed-loop real-robot deployment, NavWAM improves over planning-based world-model baselines in our evaluations while using the default policy mode without CEM-style action search. Project page: https://dachii-azm.github.io/navwam/
