Behavior of a Person-Following Robot in Pedestrian Environments: Laboratory Experimentation and Simulation

Area of Science:

• Robotics
• Human-Robot Interaction
• E-commerce Logistics

Background:

• Increasing e-commerce demands efficient last-mile delivery solutions.
• Person-following robots are emerging for delivery tasks in public areas.
• Performance in crowded, unstructured environments remains uncertain.

Purpose of the Study:

• To evaluate the performance of person-following delivery robots in dynamic pedestrian environments.
• To develop and validate a simulation model for robot performance assessment.
• To analyze robot behavior under various crowd densities.

Main Methods:

• Laboratory experiments to characterize robot operating parameters.
• Development of a robot simulation model in a pedestrian simulator.
• Calibration and validation of the simulation model with experimental data.
• Expansion of the model to include diverse pedestrian crowd scenarios.

Main Results:

• The study established a validated simulation framework for robot performance evaluation.
• Robot performance was assessed across different crowding levels.
• Empirical data informed the simulation model for realistic scenario testing.

Conclusions:

• Simulation provides a viable method for assessing robot performance in complex environments.
• Further research is needed to optimize robot navigation and interaction in crowded public spaces.
• The developed model can inform the deployment of delivery robots in real-world scenarios.