Understanding Crowd Navigation Behavior: Human-Centered Research for Social Robotics

Overview

This project examines how people navigate in crowded environments and how these behaviors can inform the development of socially aware robotic systems. My contributions focused on two key areas: identifying real-world sites for testing and refining social simulation for human-robot interactions. Through qualitative research methods, iterative design, and real-world observations, this work provides insights into human movement patterns and informs the design of robotic navigation strategies that align with social norms.

Methods & Iterative Research Design

1. Site Selection and Justification

To ensure that robotic crowd navigation strategies are tested in realistic environments, I led the effort to identify and justify real-world sites for evaluation. This involved:

• Environmental Analysis: Identifying diverse public spaces with varying crowd densities and movement complexities.

• Comparative Assessment: Evaluating potential sites based on spatial constraints, pedestrian flow, and social interaction norms.

• Field Observations: Conducting site visits to document movement patterns and crowd behaviors relevant to robotic navigation.

• Site Replication for Simulation: Collaborating with simulation teams to model these locations digitally for controlled testing scenarios.

This approach ensured that selected sites reflected real-world navigation challenges, allowing us to bridge the gap between simulation-based and field-based evaluations.

2. Refining Social Simulation

To develop more socially aware robotic behaviors, I conducted qualitative studies focused on how people move in crowds and respond to implicit social cues. This process involved:

Expressive Behavior Studies

• Interviews with Movement Experts: Collaborated with actors, puppeteers, and animators to understand how intentional movement conveys social meaning.

• Behavioral Coding & Thematic Analysis: Extracted key principles of movement that could be adapted for robotic navigation.

• Iterative Design Sessions: Worked with designers to translate human movement strategies into robotic behaviors that align with social norms.

Ethnographic Crowd Observations

• Field Research & Live Data Analysis: Studied pedestrian movement in high-density environments via first-person recordings and EarthCam live feeds.

• Micro-Interaction Analysis: Examined subtle social signals (e.g., body sways, shoulder shifts) that facilitate seamless crowd navigation.

• Behavioral Modeling Insights: Identified patterns in pedestrian behavior that inform real-time robotic decision-making.

Key Findings & Design Implications

• Social Navigation as a Nonverbal Dialogue: Pedestrians unconsciously communicate movement intentions through subtle shifts in body posture rather than explicit signals.

• Space-Holding & Adaptive Flow: People maintain dynamic "personal bubbles" that expand and contract based on crowd density, a principle robots must respect.

• Behavioral Context Matters: Robots that abruptly start or stop movement without expressive signaling risk disrupting social flow.

Impact & Next Steps

This research contributes to the design of robots that navigate public spaces in ways that feel intuitive and non-disruptive to human pedestrians. By iteratively refining our understanding of social navigation through real-world testing, simulation, and qualitative inquiry, we lay the groundwork for autonomous systems that can coexist smoothly in human environments.

Future work will focus on testing these findings in live robotic deployments and refining behavior models based on further field studies.