3D Absolute pose estimation of a staircase cleaning robot utilizing staircase geometry and minimal low-cost sensors

Abstract

Thanks to its shape-shifting abilities, the sTetris robot can climb stairs while performing the cleaning task. The sTetris robot’s overall system operation depends on localization and positioning data, which are essential for its goal of autonomously navigating multi-floor environments. The mobile robots designed to work indoors generally rely on external systems for localization information. Regretfully, this frequently requires additional hardware fixing or changes to the indoor working environment in order to achieve accurate three-dimensional (3D) position and orientation (pose) for successful operation of the mobile platform. Nonetheless, the robot can be localized on the staircase by utilizing the information of the staircase’s geometry measurements, which are known ahead of time. This article demonstrates how the known geometry of staircases and measurements from minimal number of sensors can be used to accomplish 3D pose of the robot. Experiments carried out on a real robot in an authentic indoors setting successfully demonstrate the effectiveness of the suggested approach.