Human-aware robot navigation using the social force model

Today, the long-lived dream of employing autonomous mobile robots in human environments is more close to coming into reality than ever. Most of the fundamental navigation problems are solved due to an extensive research on the field of mobile robotics. Today, there are many navigation strategies that enable robots to navigate through their surrounding environments in a collision-free manner. However, Human-Robot Interaction (HRI), suggests that, in addition to being safe, navigation of a robot must be understandable to humans, as well. Comprehensibility of the movements of a robot increases its acceptance among humans. This way, it is believed by many that, we will begin to internalize the presence of mobile robots in our daily lives. The concept of human-aware navigation refers to these shortcomings of the previous methods in this direction. Social Force Model was proposed to describe the motion of the pedestrians. We use this approach to develop a human-aware navigation module for our tour-guide robot. This model is a variant of a large-family, so-called artificial potential fields. The nature of the model makes it a good candidate for local path planning. In this work, we present the model and the differences of it from fundamental path planning algorithms. We explain the implementation of the model as a local path planner and show on simulated environment that a more socially-acceptable motion may be achieved using this model.