Cleaning robots are one of the market dominators in the commercialized robot space. So far, numerous robots have been introduced that can perform cleaning tasks in various settings, including floor, pavement, pool, lawn, windows, etc. However, none of the existing commercial cleaning robots targets the staircase, commonly found in multi-story buildings. Even though few works in the literature introduced robotic solutions for staircase cleaning, they primarily focused on cleaning the ascending staircase often, with a loose connection to access the descending staircase. In this paper, we propose a novel autonomous reconfigurable robotic platform called sTetro-D that can autonomously detect the descending staircase, approach the step, and perform area coverage in an unknown environment. The developed autonomy framework consists of two modes which are search mode and clean mode. In search mode, we implemented an RGB-D camera-based fusion technique wherein we combined the image bounding box from DCNN (Deep Convolution Neural Network) with the depth information to find the 3D first step pose that assists the robot in approaching it precisely. After the successful stair approach, the cleaning mode enables the staircase area coverage process. We described all these aspects and concluded with an experimental analysis of the proposed robotic system in a real-world scenario. The results demonstrate that the robot has a significant performance in detecting the descending staircase, staircase approach, and area coverage.

Planning the energy-efficient and collision-free paths for reconfigurable robots in complex environments is more challenging than conventional fixed-shaped robots due to their flexible degrees of freedom while navigating through tight spaces. This article presents a novel algorithm, energy-efficient batch informed trees* (BIT*) for reconfigurable robots, which incorporates BIT*, an informed, anytime sampling-based planner, with the energy-based objectives that consider the energy cost for robot’s each reconfigurable action. Moreover, it proposes to improve the direct sampling technique of informed RRT* by defining an$L^2$greedy informed setthat shrinks as a function of the state with the maximum admissible estimated cost instead of shrinking as a function of the current solution, thereby improving the convergence rate of the algorithm. Experiments were conducted on a tetromino hinged-based reconfigurable robot as a case study to validate our proposed path planning technique. The outcome of our trials shows that the proposed approach produces energy-efficient solution paths, and outperforms existing techniques on simulated and real-world experiments.