This paper presents a concept for socially aware robot navigation with a focus on avoiding the human field of view (FOV). The aim is to improve user acceptance, minimize hu- man discomfort and reduce impairment of concentration in the workplace, aspects often measured through subjective comfort and distraction metrics. For this purpose, a navigation concept is developed that integrates the human viewing direction into a costmap for navigation. Three different ways of modeling the environment around a human are implemented and then tested for comparison. The implementation for ROS2 is tested with a real robot, the Husky A200, equipped with the ZED2i camera. The concept and each variant are tested with test persons in a realistic working environment and evaluated in a user study. The results show a slight improvement in socially aware navigation by integrating the viewing areas into a costmap.

@inproceedings{WonDeSeRo25_Homeyer-Nimz,
 title = {{Socially Aware Robot Navigation: Evaluating the Impact of Human Field of View}},
 author = {Henriette Homeyer-Nimz and Hauke Petersen and Benjamin Noack},
 booktitle = {Proceedings of the Workshop on Autonomous
Delivery and Service Robot on Pedestrian and Cycle Paths},
 address = {Schkeuditz, Germany},
 month = may,
 url = {https://nbn-resolving.org/urn:nbn:de:bsz:105-qucosa2-981145},
 year = {2025}
}