WAVE-Go：面向轮腿机器人的自适应执行世界模型导航
WAVE-Go: World-Model Navigation with Adaptive Execution for Wheel-Legged Robots
- Beijing Institute of Technology（北京理工大学）
- The University of Hong Kong（香港大学）
- GAC R&D Center（广汽研究院）
- LimX Dynamics
- Beijing Academy of Artificial Intelligence (BAAI)（北京人工智能研究院）
机构由 AI 辅助整理，请以论文原文为准。
中文总结 AI 辅助
WAVE-Go提出可中断的自适应执行机制，将世界模型预测与命令执行分离，在动态环境中提升轮腿机器人导航成功率并降低碰撞率，平衡性能与规划开销。
中文摘要 AI 辅助
世界模型能够预测导航动作的后果，但预测的动作序列在执行过程中可能失效，尤其是当轮腿机器人遇到动态障碍物或改变运动模式时。我们提出WAVE-Go，一个图像目标导航框架，将世界动作预测与可中断的命令执行分离。其执行器自适应地选择动作前缀，并在更新的观测使执行失效时取消待处理命令。条件风险公式规定了在估计的累积失败预算下的前缀选择，而姿态和运动模式转换需要间隙、稳定性和任务证据检查。在报告的导航评估中，WAVE-Go实现了74.1%的分布内成功率和63.3%的动态分布外成功率，分别超过最强基线4.7和7.7个百分点，同时将碰撞从每100米4.4次减少到2.9次。与可中断的固定四命令执行相比，WAVE-Go将成功率提高了4.0个百分点，同时将重规划频率降低了51.2%，碰撞率降低了6.5%。执行消融实验还表明，运行时中断提高了成功率、碰撞率和反应延迟，但代价是额外的重规划。这些结果支持自适应、可中断的执行作为平衡导航性能和规划开销的手段。代码可在该https URL获取。
英文摘要
World models can anticipate the consequences of navigation actions, but predicted action sequences may become invalid during execution, especially when wheel-legged robots encounter dynamic obstacles or change locomotion modes. We propose WAVE-Go, an image-goal navigation framework that separates world-action prediction from interruptible command execution. Its executor adaptively selects an action prefix and cancels pending commands when updated observations invalidate execution. A conditional-risk formulation specifies prefix selection under an estimated cumulative failure budget, while posture and locomotion-mode transitions require clearance, stability, and task-evidence checks. In the reported navigation evaluation, WAVE-Go achieves 74.1% in-distribution success and 63.3% dynamic out-of-distribution success, exceeding the strongest baseline by 4.7 and 7.7 percentage points, respectively, while reducing collisions from 4.4 to 2.9 per 100 m. Compared with interruptible fixed four-command execution, WAVE-Go raises success by 4.0 percentage points while reducing replanning frequency by 51.2% and collision rate by 6.5%. Execution ablations also show that runtime interruption improves success, collision rate, and reaction latency at the cost of additional replanning. These results support adaptive, interruptible execution as a means of balancing navigation performance and planning overhead. Code is available at https://github.com/vigorlee/wave-go.
