Navigation System Development on ATRACBOT using F-SFM

Navigation systems are very important for social robots, especially autonomous robots, because they are required to be adaptive to their surrounding environment. This paper proposes a new navigation system for the Autonomous Trash Can Robot (ATRACBOT) using the Fuzzy-based Social Force Model (F-SFM). The navigation system employed in earlier stages of development depended on the Braitenberg approach, leading to the robot’s motion appearing rigid. On the other hand, the Social Force Model (SFM) employs the forces surrounding objects to engage with and evade other entities, presenting a different approach. The magnitude of the force to avoid objects is determined by the SFM parameters and the stimulus received by the robot. The Fuzzy Inference System (FIS) is used to obtain the optimal value of the gain factor, k, which is one of the SFM parameters, based on the relative distance, d, and relative angle, $\gamma$, between the obstacle and robot. The FIS will process both inputs and produce a suitable gain factor, k, so that the stimulus received can match the robot’s reactivity and responsiveness. Based on the results from a realistic 3D simulator V-Rep, the implementation of the F-SFM method can deliver the robot to the goal point without hitting obstacles and make the robot’s movements more natural compared to the Braitenberg method.