VbotNav-0: Asynchronous Cloud-Edge Dual-Policy Navigation across Indoor and Outdoor Environments

Key points

• We present VbotNav-0, a cloud-edge dual-policy framework that links these processes through asynchronous policy-to-policy conditioning.1
• The cloud policy produces subgoal guidance from visual history, navigation instructions, and route tokens; the edge policy combines available guidance with current RGB and LiDAR observations to generate local trajectories.1
• This formulation allows the local perception-action loop to continue between cloud updates.1
• A common route representation connects indoor paths and outdoor map routes, and the training recipe combines navigation demonstrations, corrective data, and grounded visual question answering (VQA).1

Sources (1)

• [1]VbotNav-0: Asynchronous Cloud-Edge Dual-Policy Navigation across Indoor and Outdoor Environments
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 10, 06:33 AM

  Continuous indoor-outdoor navigation connects destination-level reasoning with motion decisions that must respond to current observations.

  We present VbotNav-0, a cloud-edge dual-policy framework that links these processes through asynchronous policy-to-policy conditioning.

Extractive summary: sentences quoted from the sources.