Bi3: A Biplatform, Bicultural, Biperson Dataset for Social Robot Navigation

2026-05-07

AI Summary

This study addresses the challenge of enabling social robots to navigate efficiently and naturally among dense crowds while interacting with multiple people. Conducting dual-platform robotic experiments in both the United States and France, the research collected multimodal interaction data from 74 participants across five navigation algorithms. It presents the first benchmark dataset for social navigation that integrates dual robotic platforms, dual cultural contexts, and close-proximity human–robot dyadic interactions, yielding high diversity and complexity. The dataset comprises ground-truth trajectories, RGB video recordings, and subjective user evaluations, amounting to 10.5 hours of high-quality human–robot motion and interaction data. This resource provides a critical foundation for training and evaluating robot motion prediction and control strategies in densely populated environments.

Abstract

We contribute Bi3, a dataset of social robot navigation among groups of people in a constrained lab space. Compared to prior data collection efforts for social robot navigation, our dataset is unique in that it features: an original experiment design giving rise to close navigation encounters between two humans and a robot; five different navigation algorithms; two different robot platforms; a diverse participant pool of 74 people recruited from two sites in the USA and France; multimodal data streams including 10.5 hours of human and robot ground-truth motion tracks, RGB video, and user impressions over robot performance. Our analysis of the collected dataset through metrics like interaction density and human velocity suggests that Bi3 represents a benchmark of unique diversity and modeling complexity. Bi3 contributes towards understanding how humans and robots can productively mesh their activities in constrained environments, and can be a resource for training models of human motion prediction and robot control policies for navigation in densely crowded spaces.

Problem

• social robot navigation
• human-robot interaction
• crowded environments
• motion prediction
• navigation benchmark

Innovation

• social robot navigation
• multimodal dataset
• bicultural human-robot interaction
• constrained environment navigation
• heterogeneous robot platforms