This paper details an autonomous cooperative wall-building system using UAVs developed for Challenge 2 of the MBZIRC 2020 competition. The system employs scanning, RGB-D detection, precise grasping, and multi-UAV coordination to place bricks on a wall. The CTU-UPenn-NYU approach achieved the highest score in the competition by correctly placing the most bricks. Why it matters: This demonstrates advanced capabilities in robotics and autonomous systems relevant for construction and infrastructure development in challenging environments.
This paper presents two robotic systems developed for the MBZIRC 2020 competition, designed for autonomous wall construction. The systems utilize a UGV with 3D LiDAR for precise brick pose estimation and a UAV employing real-time visual servoing. The authors report results from the competition and lab experiments, discussing lessons learned from the autonomous wall-building task. Why it matters: The work highlights advancements in mobile manipulation and autonomous robotics, with potential applications in construction and infrastructure development in the region.
Team NimbRo presented their UGV solution for autonomous wall building and firefighting at the Mohamed Bin Zayed International Robotics Challenge (MBZIRC) 2020. The robot integrates a wheeled omnidirectional base, a 6 DoF manipulator arm with a magnetic gripper, a storage system, and a water spraying system. It uses 3D LiDAR, RGB, and thermal cameras to perceive the environment, pick up boxes, construct walls, and detect/extinguish fires. Why it matters: The work highlights advancements in autonomous robotics for complex tasks relevant to construction and disaster response in the UAE and globally.
In a 2018 keynote, Saudi Aramco VP Nasser Al-Nafisee recounted the rapid construction of KAUST. Al-Nafisee described building KAUST in under three years as a "mission impossible" requiring immense effort. He advised KAUST attendees to push beyond their comfort zones and adopt a "can-do attitude". Why it matters: The talk highlights the ambitious vision and rapid development that characterize Saudi Arabia's investments in research and technology.
This paper introduces a hybrid deep learning and machine learning pipeline for classifying construction and demolition waste. A dataset of 1,800 images from UAE construction sites was created, and deep features were extracted using a pre-trained Xception network. The combination of Xception features with machine learning classifiers achieved up to 99.5% accuracy, demonstrating state-of-the-art performance for debris identification.