CoHAN-Nav2

Read full text
Authors: Phani Teja Singamaneni
Cited by: 0

Abstract

CoHAN-Nav2 is an enhanced version of the Cooperative Human-Aware Navigation framework, building on the foundations of the original CoHAN system Singamaneni et al., IROS 2021. This new iteration significantly improves the stability and robustness of the navigation pipeline in dynamic human environments. A key advancement in CoHAN-Nav2 is the integration of behavior trees for context-aware mode switching, enabling more structured and scalable decision-making during navigation. The system maintains tight integration with the ROS navigation stack and leverages the Human-Aware Timed Elastic Band (HATEB) planner for trajectory generation, while the newly introduced behavior-based layer allows the robot to seamlessly adapt between social contexts and allows easier integration of new behaviors. In addition to behavior-based mode switching, CoHAN-Nav2 incorporates reasoning about invisible humans in the environment, as introduced in Singamaneni et al., IROS 2022. By fusing observed human motion with static map information, the system infers the presence of humans who may be occluded or temporarily out of sensor range. This allows the planner to proactively account for hidden human activity—enhancing safety and social compliance in complex, cluttered spaces.

Fields of Study & Subjects

Robotic Path Planning Algorithms Robotics and Sensor-Based Localization Evacuation and Crowd Dynamics

Publication details

Journal / source
Zenodo (CERN European Organization for Nuclear Research)
Publication date
2 March 2026
Language
EN