Empowering robots to navigate in a socially compliant manner is essential for the acceptance of robots moving in human-inhabited environments. Previously, roboticists have developed classical navigation systems with decades of empirical validation to achieve safety and efficiency. However, the many complex factors of social compliance make classical navigation systems hard to adapt to social situations, where no amount of tuning enables them to be both safe (people are too unpredictable) and efficient (the frozen robot problem). With recent advances in deep learning approaches, the common reaction has been to entirely discard classical navigation systems and start from scratch, building a completely new learning-based social navigation planner. In this work, we find that this reaction is unnecessarily extreme: using a large-scale real-world social navigation dataset, SCAND, we find that classical systems can be used safely and efficiently in a large number of social situations (up to 80%). We therefore ask if we can rethink this problem by leveraging the advantages of both classical and learning-based approaches. We propose a hybrid strategy in which we learn to switch between a classical geometric planner and a data-driven method. Our experiments on both SCAND and two physical robots show that the hybrid planner can achieve better social compliance in terms of a variety of metrics, compared to using either the classical or learning-based approach alone.

翻译：赋予机器人在人类居住环境中的社交合规导航能力，对于提升其被接纳程度至关重要。此前，机器人专家开发了经过数十年实证验证的经典导航系统，以实现安全性和效率。然而，社交合规性涉及的诸多复杂因素使经典导航系统难以适应社交场景——无论怎样调整参数，都无法同时保证安全性（人类行为过于不可预测）和效率（“冻结机器人”问题）。随着深度学习方法的近期进展，普遍反应是完全摒弃经典导航系统，从零开始构建全新的基于学习的社交导航规划器。本研究发现这种反应过于极端：通过大规模真实社交导航数据集SCAND，我们发现经典系统在高达80%的社交场景中仍能安全高效地运行。因此，我们重新审视该问题，试图融合经典方法与学习方法的优势，提出一种混合策略：学习在经典几何规划器与数据驱动方法之间进行切换。基于SCAND数据集及两台实体机器人的实验表明，与单独使用经典或学习方法相比，混合规划器在多项指标上均实现了更优的社交合规性。