Planning Socially Expressive Mobile Robot Trajectories

Area of Science:

• Robotics
• Human-Robot Interaction
• Artificial Intelligence

Background:

• Mobile robots navigating human environments require understanding human social perceptions of robot motion.
• Existing social navigation algorithms lack configurability to control specific social interpretations.
• Human perception of robot motion is linked to social attitudes and intentions.

Purpose of the Study:

• To identify key robot motion features influencing human social perception.
• To develop a trajectory planning method for controlling social perceptions in robot navigation.
• To enhance robot acceptability by tailoring motion to context and role.

Main Methods:

• Developed logistic regression models based on human perception experiments and robot velocity profiles.
• Formulated a constrained optimization problem for trajectory planning.
• Introduced novel constraints to shape robot trajectories for desired social perceptions.

Main Results:

• Established that specific trajectory features significantly impact human social perception of robots.
• Demonstrated the algorithm's ability to accurately modify trajectory features based on selected constraints.
• Showcased consistent application of subtle motion variations to induce specific social perceptions.

Conclusions:

• Robot motion significantly influences human social perception, affecting acceptability.
• The proposed trajectory planning method allows for controlled generation of social perceptions.
• This approach offers a tool to enhance robot integration into human environments by tailoring motion.