Socially Aware Navigation: A Non-linear Multi-objective Optimization Approach

Mobile robots are increasingly populating homes, hospitals, shopping malls, factory floors, and other human environments. Human society has social norms that people mutually accept; obeying these is an essential signal someone participating socially with respect to the rest of population. For be compatible humans, it crucial for obey norms. In prior work, we demonstrated a Socially-Aware Navigation (SAN) planner, based on Pareto Concavity Elimination Transformation (PaCcET), in hallway scenario, optimizing two objectives so robot does not invade personal space people. This article extends our PaCcET-based SAN planner multiple scenarios more than objectives. We modified Robot Operating System’s (ROS) navigation stack include PaCcET local planning task. show approach can accommodate Human-Robot Interaction (HRI) scenarios. Using proposed approach, achieved successful HRI such as interactions, art gallery, waiting queue, interacting group. implemented method simulated PR2 2D simulator (Stage) pioneer-3DX mobile real-world validate all A comprehensive set experiments shows handle interaction both holonomic non-holonomic robots; hence, viable option Unified (USAN).