5 May 2021 in IEEE/ASME Transactions on Mechatronics
Façade cleaning of high-rise buildings requires automation due to the risk and inefficiency of a fall accident for human workers. Recently, due to the need for automation, various types of façade cleaning robots have been developed. In our previous study, the gondola-mounted type cleaning robot was developed, and it showed several issues. In this article, we propose an improved design feature to solve the issues. First, we modularized the robot design for lightweight and maintainability. Second, we implemented passive obstacle overcoming mechanism with tri-star wheel and compliant manipulator for minimizing uncleaned area near obstacles. Third, we developed the position sensing device for measurement and compensation of lateral disturbance occurred by swinging gondola. To verify the cleaning performance of the robot, experiments were conducted with different scenarios of cleaning general plane surface, obstacle overcoming, and lateral disturbance applied and compensated. Experimental results of both lab test and field-test performed on a real-world building were analyzed and summarized in this article.
4 May 2020 in IEEE/ASME Transactions on Mechatronics
The number of high-rise buildings is increasing worldwide, and the demand for window maintenance and cleaning has increased accordingly. However, there is an increased desire to substitute robots with human workers because of the occurrence of many accidents. Although many window cleaning robots have been developed, almost all the robots cannot passively pass obstacles on the outer walls of buildings. In this study, a window cleaning robot is provided with compliant adaptation capabilities by adding series elastic actuators (SEAs) and tri-star wheels to an older window cleaning robot. This robot has a two-degree-of-freedom manipulator, and the robot can passively overcome obstacles without any need for control. An optimization is performed for the manipulator, and experiments are performed with the manipulator to compare the respective abilities to overcome obstacles using parameters before and after the optimization. In addition, a cleaning module is constructed by scaling down the previous model. The performance of the downscaled cleaning module is validated using a window cleaning experiment.
30 September 2020 in IEEE Access
In recent years, as number of new building getting larger, there has been an increased interest in the cleaning of exterior walls. Accordingly, there is a growing interest in automatic cleaning robots that move around the outer building fagade. These robots are also required to apply different cleaning methods to remove various contaminants on the outer wall of the building. However, current surface contaminant detection systems can either detect only a single type of contaminant, or are not compact enough for installation on mobile platforms that move around the outer fagade. As cleaning workers are able to distinguish various contaminants with the naked eye, we aim to solve this problem by developing a machine-vision system using convolutional neural networks (CNNs) and image processing methods. As it is a compact system that uses only a camera to take pictures and a processor to process the images, it is suitable for applications involving mobile platforms. Object-type contaminants such as avian feces are handled by the YOLOv3 module using the object-detection algorithm. Area-type contaminants such as rusty stains are processed using the color-detection module using the HSV color space, median filter, and flood fill algorithm. Particle-type contaminants such as dust are handled by the grayscale module, converting images to grayscale images and then comparing the average brightness with a reference that is provided in advance. This proposed machine vision system will detect objects, areas, and particle-type contaminants with a single image and some reference images provided in advance.
31 October 2019 in IEEE Access
As the demand and market for building maintenance are increasing, automated building façade cleaning has become essential. Robots are replacing human workers because cleaning work on high-rise buildings using gondolas can be dangerous. Several façade cleaning robots have been developed for climbing, and practical knowledge to clean the façade is being adopted in their cleaning devices. In this study, a passive linkage suspension mechanism and tri-star wheels are applied to solve the problems of unclean zones due to failures during overcoming obstacles and the problems through the use of additional actuators. Various mechanism models have been introduced and their performances have been compared based on dynamic simulation considering obstacle encounters.
6 January 2020 in IEEE Access
This paper proposes a three-degree-of-freedom manifold composed of three linear actuators. The proposed mechanism consists of a workspace suitable for facade cleaning and can compensate for the horizontal position from disturbances in a gondola-based exterior-wall cleaning. We design a cleaning manipulator that can ensure a constant cleaning area by compensating for the disturbance in each direction. The position, velocity kinematic, and Jacobian-based singularity analysis are presented, and kinematic variables are defined to extend a singularity-free workspace. In addition, optimization is performed based on an index that demonstrates the mechanical properties of the manipulator. The result shows how the manipulator compensates for the disturbances as well as the features of the optimization model. This study can be applied to robot manipulators for facade cleaning in the future.
25 December 2019 in IEEE Access
As the number of high-rise buildings is increasing, more methods of exterior-wall cleaning are being developed. There are a few models based on artificial intelligence that determine the type and level of contamination primarily by moving the cleaning area. In this study, we propose an system using YOLOv3 algorithm, color-detection, to install on façade cleaning robot and brightness-discrimination. There are three types of contaminant-detection parameters: size, color, and brightness, and these parameters are subjected to a robust optimization process to maintain a constant detection rate under different conditions. The three parameters are determined via Taguchi method with signal to noise ratio and noise factors. An environment for algorithm testing is established, and artificial contamination is implemented on the specimen. A field test with the detection algorithm shall be performed in the near future.