Article
Robotics
Kateryna Zorina, Justin Carpentier, Josef Sivic, Vladimir Petrik
Summary: Seamlessly integrating robots into human environments requires the robots to learn how to use human tools. This work introduces an automated approach that learns tool manipulation strategies from Youtube videos instead of expert demonstrations. The approach includes an alignment procedure to align the simulated environment with the real-world scene in the video, and combines reinforcement learning and optimization to learn control policies. The effectiveness of the proposed approach is demonstrated in simulation and with a real Franka Emika Panda robot demonstration.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2022)
Article
Robotics
Takayuki Osa, Naoto Osajima, Masanori Aizawa, Tatsuya Harada
Summary: This article investigates the use of reinforcement learning to automate construction machines and proposes two techniques to improve performance. The study shows that this approach is more effective in excavation tasks compared to existing meta-learning and domain adaptation methods.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2023)
Article
Automation & Control Systems
Daniel Sanchez-Martinez, Carlos A. Jara, Francisco Gomez-Donoso
Summary: Nowadays, there are many industrial processes that involve tedious and repetitive tasks, often with dangerous materials or machinery. This paper presents an automatic and innovative collaborative robotic system that can handle the demoulding task in toy doll manufacturing. The system uses a vision-based algorithm and a custom gripper integrated into a UR10e robot to detect and extract the pieces from the mold.
INTERNATIONAL JOURNAL OF ADVANCED MANUFACTURING TECHNOLOGY
(2023)
Article
Automation & Control Systems
Gwendolyn Foo, Sami Kara, Maurice Pagnucco
Summary: This paper presents an effective learning framework for robotic disassembly of LCD monitors, which significantly improves the successful part identification rate from 11% to 87% after being trained on past disassembly experience.
IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING
(2023)
Article
Automation & Control Systems
Gwendolyn Foo, Sami Kara, Maurice Pagnucco
Summary: Disassembly is a crucial step in the treatment of waste electrical and electronic equipment (WEEE) to prevent environmental damage. Automation and robotics face challenges due to the vast variations in WEEE, prompting the need for flexible intelligence in handling different situations. This paper introduces an ontology-based cognitive method for disassembly actions, focusing on LCD monitors, to address uncertainties and improve robotic disassembly processes.
INTERNATIONAL JOURNAL OF AUTOMATION TECHNOLOGY
(2021)
Article
Robotics
Hongzhi Du, Yanyan Li, Yanbiao Sun, Jigui Zhu, Federico Tombari
Summary: This research delves into the cost aggregation strategy and GPU memory consumption issues in learning-based stereo matching tasks, proposing a novel recurrent cost aggregation strategy and Stacked Recurrent Hourglass (SRH) module that successfully reduce GPU memory consumption. Meanwhile, performance is enhanced through multi-scale information processing in textureless areas.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2021)
Article
Robotics
Nicholas S. Selby, H. Harry Asada
Summary: This paper presents a method for learning effective and causal observable functions for low-order lifting linearization of nonlinear controlled systems from data, and eliminates anti-causal components of the observables to lift the system.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2021)
Article
Engineering, Industrial
Jan F. Hellmuth, Nicholas M. DiFilippo, Musa K. Jouaneh
Summary: Electric vehicles offer an environmentally friendly transportation solution, with high sales predicted in the future. The most costly components of these vehicles are their batteries, which must be recycled after use. This study presents a methodology to assess the automation potential of disassembling EV batteries, using criteria to determine which steps should be automated and which should be done manually.
JOURNAL OF MANUFACTURING SYSTEMS
(2021)
Article
Biochemical Research Methods
Eladio Rodriguez-Diaz, Samer Kaanan, Christopher Vanley, Tauseef Qureshi, Irving J. Bigio
Summary: The study examines the feasibility of using ESS coupled with machine learning to accurately distinguish between different tissue types in lung excision surgeries, demonstrating potential for improving accuracy in targeting pulmonary lesions.
JOURNAL OF BIOPHOTONICS
(2021)
Article
Ergonomics
Jennifer M. Lincoln, K. C. Elliott
Summary: Various factors are driving the development of robotics and automation in the agriculture industry, providing occupational safety and health researchers an opportunity to mitigate risks and benefit agriculture workers' health and safety.
JOURNAL OF SAFETY RESEARCH
(2023)
Article
Computer Science, Information Systems
Jingjing Wan, Lechen Sun, Tianhao Du
Summary: Compared with rigid robots, soft robots have the advantages of strong environmental adaptability, good human-robot interaction, and simple control systems. This study focuses on the development and application of three kinds of pneumatic soft actuators. The experimental testing and simulation show that these actuators have excellent performance, fast response, high stability, and large driving force. The research also designs a soft robot for pipe crawling and a soft robotic arm, demonstrating their potential in various industrial applications.
Article
Automation & Control Systems
Antonio Serrano-Munoz, Nestor Arana-Arexolaleiba, Dimitrios Chrysostomou, Simon Bogh
Summary: This study presents the state of the art of contact-rich disassembly using reinforcement learning algorithms and investigates the generalisation capabilities of object extraction skills. The results show that reinforcement learning agents can generalise contact-rich extraction skills.
INTERNATIONAL JOURNAL OF ADVANCED MANUFACTURING TECHNOLOGY
(2023)
Article
Automation & Control Systems
Amit Bhaya
Summary: Rio de Janeiro is a popular tourist destination and home to the Federal University of Rio de Janeiro (UFRJ). The UFRJ's Graduate School of Engineering, specifically the Control, Automation, and Robotics (CAR) group in the Electrical Engineering Department, is known for its excellence in engineering research. The CAR group will host the 2025 IEEE Conference on Decision and Control, with Joao Carlos Basilio as general cochair.
IEEE CONTROL SYSTEMS MAGAZINE
(2023)
Article
Engineering, Chemical
Yuhan Wang, Chong Shen, Min Qiu, Minjing Shang, Yuanhai Su
Summary: In this work, we solved the partial differential equation of the one-dimensional axial diffusion model in the open-source platform FEniCS to understand the impact of axial dispersion on reaction yield-to-time profiles. Additionally, we developed an automatic platform consisting of a photomicroreactor, controlled pumps, UV-LED light source, absorbance analytical unit, and Raspberry Pi controller. This platform allowed for steady-state feeding and sampling functions and was validated using model reactions to study reaction mechanisms and optimize processes using genetic algorithm based symbolic regression.
Article
Computer Science, Artificial Intelligence
Marco Wurster, Marius Michel, Marvin Carl May, Andreas Kuhnle, Nicole Stricker, Gisela Lanza
Summary: Remanufacturing involves disassembling and reassembling used products to conserve resources and reduce emissions. Disassembly is a new problem in production planning and control, as it faces uncertainty in the condition of returned products which leads to volatility in remanufacturing systems. Advances in robotics and AI are enabling the automation of disassembly with autonomous workstations, but there is a need to address the risk of operational failures through a condition-based control approach.
JOURNAL OF INTELLIGENT MANUFACTURING
(2022)
Article
Computer Science, Information Systems
Mohsen Rezvani, David Rajaratnam, Aleksandar Ignjatovic, Maurice Pagnucco, Sanjay Jha
INTERNATIONAL JOURNAL OF INFORMATION SECURITY
(2019)
Article
Computer Science, Interdisciplinary Applications
Wei Hua Chen, Gwendolyn Foo, Sami Kara, Maurice Pagnucco
Summary: The study introduces a robotic disassembly system aimed at addressing product variation and end-of-life product and condition information uncertainties, capable of dismantling LCD screens with user intervention and flexible planning. By combining specific product information with background knowledge, the system is able to assess and select desirable disassembly actions effectively.
ROBOTICS AND COMPUTER-INTEGRATED MANUFACTURING
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Yiwen Xu, Maurice Pagnucco, Yang Song
Summary: This paper proposes a Decoupled High-frequency semantic Guidance-based GAN (DHG-GAN) for diverse image outpainting, which aims to restore large missing regions surrounding a known region while generating multiple plausible results. Experimental results demonstrate that the proposed method outperforms existing approaches on CelebA-HQ, Place2, and Oxford Flower102 datasets.
COMPUTER VISION - ACCV 2022, PT VII
(2023)
Proceedings Paper
Computer Science, Artificial Intelligence
Cong Cong, Yixing Yang, Sidong Liu, Maurice Pagnucco, Antonio Di Ieva, Shlomo Berkovsky, Yang Song
Summary: This paper proposes a novel unified feature and classifier learning framework for imbalanced medical image datasets. The model is equipped with an adaptive unified contrastive (AduC) loss to progressively adapt the model learning process. Experimental results show that the proposed method can significantly improve the classification accuracy and F1-score for medical image classification.
MACHINE LEARNING IN MEDICAL IMAGING, MLMI 2022
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Lihuan Li, Maurice Pagnucco, Yang Song
Summary: Pedestrian trajectory prediction is crucial and challenging for applications like autonomous driving and robotic motion planning. Existing methods often neglect the smoothness and temporal consistency of predictions. This study proposes a model that can generate multiple paths, correct inconsistent trajectories, and achieve state-of-the-art performance in multi-future prediction.
2022 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2022)
(2022)
Proceedings Paper
Imaging Science & Photographic Technology
Kunzi Xie, Yixing Yang, Maurice Pagnucco, Yang Song
Summary: This paper proposes a new unsupervised method for electron microscope image registration, which integrates a Cascaded LST-UNet framework with Swin Transformer to achieve better alignment.
MEDICAL IMAGE COMPUTING AND COMPUTER ASSISTED INTERVENTION, MICCAI 2022, PT VI
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Maurice Pagnucco, David Rajaratnam, Raynaldio Limarga, Abhaya Nayak, Yang Song
Summary: The paper aims to integrate ethical principles into autonomous machines and provides a method for reasoning about actions to guide moral and ethical choices. Reasoning about knowledge can be used to guide ethical decision-making in dynamic scenarios for autonomous machines.
AIES '21: PROCEEDINGS OF THE 2021 AAAI/ACM CONFERENCE ON AI, ETHICS, AND SOCIETY
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Ruoyu Guo, Maurice Pagnucco, Yang Song
Summary: A weakly supervised learning method for nuclei segmentation that only requires annotation of the nuclear centroid was proposed, which utilized a mask-guided attention auxiliary network and Confident Learning to improve pixel-level labels, achieving highly competitive performance on two public datasets for cell nuclei segmentation.
MEDICAL IMAGE COMPUTING AND COMPUTER ASSISTED INTERVENTION - MICCAI 2021, PT II
(2021)
Proceedings Paper
Acoustics
Cong Cong, Sidong Liu, Antonio Di Ieva, Maurice Pagnucco, Shlomo Berkovsky, Yang Song
Summary: This study introduces a method that combines semi-supervised learning with GAN to integrate source domain images in the learning of stain normalization without the need for corresponding ground truth data. Our approach demonstrates highly effective performance on two classification tasks for brain and breast cancers.
MEDICAL IMAGE COMPUTING AND COMPUTER ASSISTED INTERVENTION - MICCAI 2021, PT VIII
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Han Lin, Maurice Pagnucco, Yang Song
Summary: Deep-learning based generative models have shown great success in various image processing tasks, capable of regenerating semantically coherent images from limited input information, even when structural features are missing.
2021 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION WORKSHOPS, CVPRW 2021
(2021)
Proceedings Paper
Automation & Control Systems
Cong Cong, Zhichao Yang, Yang Song, Maurice Pagnucco
16TH IEEE INTERNATIONAL CONFERENCE ON CONTROL, AUTOMATION, ROBOTICS AND VISION (ICARCV 2020)
(2020)
Proceedings Paper
Computer Science, Artificial Intelligence
Anna Trofimova, Timothy Wiley, Maurice Pagnucco, Mari Velonaki
PROCEEDINGS OF THE 31ST AUSTRALIAN CONFERENCE ON HUMAN-COMPUTER-INTERACTION (OZCHI'19)
(2020)
Proceedings Paper
Computer Science, Theory & Methods
Christoph Schwering, Maurice Pagnucco
AAMAS '19: PROCEEDINGS OF THE 18TH INTERNATIONAL CONFERENCE ON AUTONOMOUS AGENTS AND MULTIAGENT SYSTEMS
(2019)
Article
Computer Science, Artificial Intelligence
Christoph Schwering, Gerhard Lakemeyer, Maurice Pagnucco
ARTIFICIAL INTELLIGENCE
(2017)
Article
Logic
Zhiqiang Zhuang, Maurice Pagnucco, Yan Zhang
JOURNAL OF PHILOSOPHICAL LOGIC
(2017)
Article
Computer Science, Interdisciplinary Applications
Shenglin Wang, Jingqiong Zhang, Peng Wang, James Law, Radu Calinescu, Lyudmila Mihaylova
Summary: In Industry 5.0, Digital Twins provide flexibility and efficiency for smart manufacturing. Deep learning techniques are used to enhance the Digital Twin framework, enabling the detection and classification of human operators and robots during the manufacturing process. The framework shows promising results in accurately detecting and classifying actions of human operators and robots in various scenarios.
ROBOTICS AND COMPUTER-INTEGRATED MANUFACTURING
(2024)
Article
Computer Science, Interdisciplinary Applications
Yi Liu, Junpeng Qiu, Jincheng Wang, Junhe Lian, Zeran Hou, Junying Min
Summary: In this study, a double-sided robotic roller forming process was developed to form ultrahigh strength steels to thin-walled profiles. Synchronized laser heating and iterative path compensation method were used to reduce forming forces and achieve high-precision forming.
ROBOTICS AND COMPUTER-INTEGRATED MANUFACTURING
(2024)
Article
Computer Science, Interdisciplinary Applications
Zequn Zhang, Yuchen Ji, Dunbing Tang, Jie Chen, Changchun Liu
Summary: This paper proposes a digital twin system for human-robot collaboration (HRC) that overcomes the limitations of current methods and improves the overall performance. The system includes a human mesh recovery algorithm and uncertainty estimation to enhance the system's capabilities. Experimental results demonstrate the superiority of the proposed methods over baseline methods. The feasibility and effectiveness of the HRC system are validated through a case study involving component assembly.
ROBOTICS AND COMPUTER-INTEGRATED MANUFACTURING
(2024)
Article
Computer Science, Interdisciplinary Applications
Junmin Park, Taehoon Kim, Chengyan Gu, Yun Kang, Joono Cheong
Summary: This paper proposes a highly reliable and accurate collision estimator for robot manipulators in human-robot collaborative environments using the Bayesian approach. By assuming robot collisions as dynamic Markov processes, the estimator can integrate prior beliefs and measurements to produce current beliefs in a recursive form. The method achieves compelling performance in collision estimation with high accuracy and no false alarms.
ROBOTICS AND COMPUTER-INTEGRATED MANUFACTURING
(2024)
Article
Computer Science, Interdisciplinary Applications
Meng Wang, Kaixuan Chen, Panfeng Wang, Yimin Song, Tao Sun
Summary: In this study, a novel teleoperation machining mode and control strategy were proposed to improve efficiency and accuracy in small batch production of large casting parts. By using variable motion mapping and elastic compensation, constant cutting force was achieved, and the workpiece was protected by employing forbidden virtual fixtures and movement constraints on the slave robot.
ROBOTICS AND COMPUTER-INTEGRATED MANUFACTURING
(2024)
Article
Computer Science, Interdisciplinary Applications
Zhaoyu Li, Dong He, Xiangyu Li, Xiaoke Deng, Pengcheng Hu, Jiancheng Hao, Yue Hou, Hongyu Yu, Kai Tang
Summary: This paper presents a novel algorithm for planning a five-axis inspection path for arbitrary freeform surfaces. By converting the inspection path planning problem into a set-covering problem, the algorithm generates a near-minimum set of inspection paths that satisfy necessary constraints. Both computer simulation and physical inspection experiments confirm the effectiveness and advantages of the proposed method.
ROBOTICS AND COMPUTER-INTEGRATED MANUFACTURING
(2024)
Article
Computer Science, Interdisciplinary Applications
Hang Zhang, Wenhu Wang, Shusheng Zhang, Yajun Zhang, Jingtao Zhou, Zhen Wang, Bo Huang, Rui Huang
Summary: This paper introduces a novel framework based on deep reinforcement learning for generating machining process routes for designated parts. The framework utilizes graph representations of parts and employs convolutional graph neural networks for effective processing. Experimental results demonstrate the ability of the proposed method to generate efficient machining process routes and overcome limitations of traditional methods.
ROBOTICS AND COMPUTER-INTEGRATED MANUFACTURING
(2024)
Article
Computer Science, Interdisciplinary Applications
Regina Kyung-Jin Lee, Hao Zheng, Yuqian Lu
Summary: Future manufacturing will witness a shift towards collaboration and compassion in human-robot relationships. To enable seamless knowledge transfer, a unified knowledge representation system that can be shared by humans and robots is essential. The Human-Robot Shared Assembly Taxonomy (HR-SAT) proposed in this study allows comprehensive assembly tasks to be represented as a knowledge graph that is understandable by both humans and robots. HR-SAT incorporates rich assembly information and has diverse applications in process planning, quality checking, and human-robot collaboration.
ROBOTICS AND COMPUTER-INTEGRATED MANUFACTURING
(2024)
Article
Computer Science, Interdisciplinary Applications
Jianhui He, Lefeng Gu, Guilin Yang, Yiyang Feng, Silu Chen, Zaojun Fang
Summary: This paper presents a new modular kinematic error model for collaborative robots and proposes a portable self-calibration device to improve their positioning accuracy.
ROBOTICS AND COMPUTER-INTEGRATED MANUFACTURING
(2024)
Article
Computer Science, Interdisciplinary Applications
Hongwei Sun, Jixiang Yang, Han Ding
Summary: This paper proposes an asymmetrical FIR filter-based tool path smoothing algorithm to fully utilize the joint drive capability of robot manipulators. The algorithm considers the pose-dependent dynamics and constraints of the robot and improves motion efficiency by over 10% compared to traditional methods.
ROBOTICS AND COMPUTER-INTEGRATED MANUFACTURING
(2024)
Article
Computer Science, Interdisciplinary Applications
Dongsheng Ge, Huan Zhao, Yiwei Wang, Dianxi Li, Xiangfei Li, Han Ding
Summary: This paper focuses on learning a stable force control policy from human demonstration during contact transients. Based on the analysis of human demonstration data, a novel human-inspired force control strategy called compliant dynamical system (CDS) is proposed. The effectiveness of the proposed method is validated through simulation and real-world experiments.
ROBOTICS AND COMPUTER-INTEGRATED MANUFACTURING
(2024)
Article
Computer Science, Interdisciplinary Applications
Xuepeng Huang, Zhenzhong Wang, Lucheng Li, Qi Luo
Summary: This study models the stiffness of a robot and modifies the tool influence function (TIF) with the Preston equation in order to achieve uniform surface quality in robotic bonnet polishing (RBP) of optical components. Experimental results validate the accuracy of the modified model.
ROBOTICS AND COMPUTER-INTEGRATED MANUFACTURING
(2024)
Article
Computer Science, Interdisciplinary Applications
Mario D. Fiore, Felix Allmendinger, Ciro Natale
Summary: This paper presents a constraint-based programming framework for task specification and motion optimization. The framework can handle constraints on robot joint and Cartesian coordinates, as well as time dependency. It also compares with existing methods and provides numerical support through illustrative examples and case studies.
ROBOTICS AND COMPUTER-INTEGRATED MANUFACTURING
(2024)
Article
Computer Science, Interdisciplinary Applications
Yongxue Chen, Yaoan Lu, Ye Ding
Summary: This paper presents an optimization method for directly generating a six-degree-of-freedom toolpath for robotic flank milling. By optimizing the smoothness of the toolpath and the stiffness of the robot, the efficiency, accuracy, and finish of the machining are improved.
ROBOTICS AND COMPUTER-INTEGRATED MANUFACTURING
(2024)
Article
Computer Science, Interdisciplinary Applications
Chungang Zhuang, Haoyu Wang, Han Ding
Summary: This article proposes an end-to-end pipeline for synchronously regressing potential object poses from an unsegmented point cloud. It extracts point pair features and uses a voting architecture for instance feature extraction, along with a 3D heatmap for clustering votes and generating center seeds. An attention voting module is also employed to adaptively fuse point-wise features into instance-wise features. The network demonstrates robustness and improved performance in pose estimation.
ROBOTICS AND COMPUTER-INTEGRATED MANUFACTURING
(2024)
