Social Robot Navigation
Overview
Social robot navigation studies how mobile robots can reach goals among people without treating them as simple moving obstacles, accounting for personal space, intent, group behavior, and perceived comfort. Common approaches combine pedestrian-motion prediction with social-force or proxemics costs, model-predictive control, reinforcement learning, imitation learning, and increasingly vision- or language-model reasoning linked to conventional planners and safety checks. Researchers evaluate these methods in crowd simulations and real-world trials, using measures such as collisions, task success, personal-space violations, pedestrian delay, and human ratings; datasets and benchmarks are expanding across settings, cultures, and robot types. A central challenge is balancing safety, efficiency, and socially appropriate behavior while making high-level semantic reasoning reliable and fast enough for control, and ensuring evaluations reflect diverse people and real interactions.
Latest papers
81 papers15 papers in the last four weeks, up 150% on the four weeks before. 0.1% of all new papers.