Deep-Learning-Based Context-Aware Multi-Level Information Fusion Systems for Indoor Mobile Robots Safe Navigation.

Hazardous object detection (escalators, stairs, glass doors, etc.) and avoidance are critical functional safety modules for autonomous mobile cleaning robots. Conventional object detectors have less accuracy for detecting low-feature hazardous objects and have miss detection, and the false classification ratio is high when the object is under occlusion. Miss detection or false classification of hazardous objects poses an operational safety issue for mobile robots. This work presents a deep-learning-based context-aware multi-level information fusion framework for autonomous mobile cleaning robots to detect and avoid hazardous objects with a higher confidence level, even if the object is under occlusion. First, the image-level-contextual-encoding module was proposed and incorporated with the Faster RCNN ResNet 50 object detector model to improve the low-featured and occluded hazardous object detection in an indoor environment. Further, a safe-distance-estimation function was proposed to avoid hazardous objects. It computes the distance of the hazardous object from the robot's position and steers the robot into a safer zone using detection results and object depth data. The proposed framework was trained with a custom image dataset using fine-tuning techniques and tested in real-time with an in-house-developed mobile cleaning robot, BELUGA. The experimental results show that the proposed algorithm detected the low-featured and occluded hazardous object with a higher confidence level than the conventional object detector and scored an average detection accuracy of 88.71%.

Coverage Path Planning Using Reinforcement Learning-Based TSP for hTetran-A Polyabolo-Inspired Self-Reconfigurable Tiling Robot.

One of the critical challenges in deploying the cleaning robots is the completion of covering the entire area. Current tiling robots for area coverage have fixed forms and are limited to cleaning only certain areas. The reconfigurable system is the creative answer to such an optimal coverage problem. The tiling robot's goal enables the complete coverage of the entire area by reconfiguring to different shapes according to the area's needs. In the particular sequencing of navigation, it is essential to have a structure that allows the robot to extend the coverage range while saving energy usage during navigation. This implies that the robot is able to cover larger areas entirely with the least required actions. This paper presents a complete path planning (CPP) for hTetran, a polyabolo tiled robot, based on a TSP-based reinforcement learning optimization. This structure simultaneously produces robot shapes and sequential trajectories whilst maximizing the reward of the trained reinforcement learning (RL) model within the predefined polyabolo-based tileset. To this end, a reinforcement learning-based travel sales problem (TSP) with proximal policy optimization (PPO) algorithm was trained using the complementary learning computation of the TSP sequencing. The reconstructive results of the proposed RL-TSP-based CPP for hTetran were compared in terms of energy and time spent with the conventional tiled hypothetical models that incorporate TSP solved through an evolutionary based ant colony optimization (ACO) approach. The CPP demonstrates an ability to generate an ideal Pareto optima trajectory that enhances the robot's navigation inside the real environment with the least energy and time spent in the company of conventional techniques.

Vision Based Wall Following Framework: A Case Study With HSR Robot for Cleaning Application.

Periodic cleaning of all frequently touched social areas such as walls, doors, locks, handles, windows has become the first line of defense against all infectious diseases. Among those, cleaning of large wall areas manually is always tedious, time-consuming, and astounding task. Although numerous cleaning companies are interested in deploying robotic cleaning solutions, they are mostly not addressing wall cleaning. To this end, we are proposing a new vision-based wall following framework that acts as an add-on for any professional robotic platform to perform wall cleaning. The proposed framework uses Deep Learning (DL) framework to visually detect, classify, and segment the wall/floor surface and instructs the robot to wall follow to execute the cleaning task. Also, we summarized the system architecture of Toyota Human Support Robot (HSR), which has been used as our testing platform. We evaluated the performance of the proposed framework on HSR robot under various defined scenarios. Our experimental results indicate that the proposed framework could successfully classify and segment the wall/floor surface and also detect the obstacle on wall and floor with high detection accuracy and demonstrates a robust behavior of wall following.

A Self-Organizing Fuzzy Logic Classifier for Benchmarking Robot-Aided Blasting of Ship Hulls.

Regular dry dock maintenance work on ship hulls is essential for maintaining the efficiency and sustainability of the shipping industry. Hydro blasting is one of the major processes of dry dock maintenance work, where human labor is extensively used. The conventional methods of maintenance work suffer from many shortcomings, and hence robotized solutions have been developed. This paper proposes a novel robotic system that can synthesize a benchmarking map for a previously blasted ship hull. A Self-Organizing Fuzzy logic (SOF) classifier has been developed to benchmark the blasting quality of a ship hull similar to blasting quality categorization done by human experts. Hornbill, a multipurpose inspection and maintenance robot intended for hydro blasting, benchmarking, and painting, has been developed by integrating the proposed SOF classifier. Moreover, an integrated system solution has been developed to improve dry dock maintenance of ship hulls. The proposed SOF classifier can achieve a mean accuracy of 0.9942 with an execution time of 8.42 µ s. Realtime experimenting with the proposed robotic system has been conducted on a ship hull. This experiment confirms the ability of the proposed robotic system in synthesizing a benchmarking map that reveals the benchmarking quality of different areas of a previously blasted ship hull. This sort of a benchmarking map would be useful for ensuring the blasting quality as well as performing efficient spot wise reblasting before the painting. Therefore, the proposed robotic system could be utilized for improving the efficiency and quality of hydro blasting work on the ship hull maintenance industry.