Article
Computer Science, Artificial Intelligence
Hsien-Pin Hsu
Summary: This paper proposes a two-stage framework to solve the three essential problems in printed circuit board assembly and conducts experiments to compare the effectiveness of different metaheuristics. The results show that the hybridization of IFA with DP performs the best in terms of total assembly time.
ADVANCED ENGINEERING INFORMATICS
(2022)
Article
Chemistry, Multidisciplinary
Utku Emre Ali, He Yang, Vladislav Khayrudinov, Gaurav Modi, Zengguang Cheng, Ritesh Agarwal, Harri Lipsanen, Harish Bhaskaran
Summary: This article introduces a technique for reliably transferring individual nanowires onto various platforms and demonstrates their applications on different structures. The versatility of this technique enables easy integration of nanowires into previously seen cumbersome or impractical applications, including TEM studies and in situ electrical, optical, and mechanical characterization.
Article
Engineering, Manufacturing
Yuqiao Cen, Jingxi He, Daehan Won
Summary: The study investigates component pick-and-place (P&P) defect patterns for different root causes using automated optical inspection data, and develops a root cause identification model with machine learning. Experimental results show that wrong nozzle size can increase component placement offset and drop probability, while nozzle pick-up position affects rotated placement offset. Machine learning methods are used to trace back these root causes.
SOLDERING & SURFACE MOUNT TECHNOLOGY
(2022)
Article
Computer Science, Information Systems
Fan Hong, Donavan Wei Liang Tay, Alfred Ang
Summary: This paper describes the development of an intelligent object detection and picking system based on MobileNet, which is integrated into a six-axis robotic arm. Experimental results show that the MobileNet model achieves an accuracy of 91%, a significant improvement compared to the original sequential model.
Article
Engineering, Biomedical
Claudia D'Ettorre, Agostino Stilli, George Dwyer, Maxine Tran, Danail Stoyanov
Summary: In this study, a novel software architecture was introduced to support hardware soft robotic rail for intra-operative ultrasound imaging automation. Preliminary results indicate the potential of the proposed semi-autonomous approach in reducing the time needed to complete ultrasound scans.
INTERNATIONAL JOURNAL OF COMPUTER ASSISTED RADIOLOGY AND SURGERY
(2021)
Article
Engineering, Manufacturing
Jingxi He, Yuqiao Cen, Shrouq Alelaumi, Daehan Won
Summary: This research proposes a novel AI-based framework that identifies the optimal placement position to minimize post-reflow misalignment of mini-scale passive components. The framework adapts to different solder paste printing locations and optimizes the model parameters to improve placement accuracy.
IEEE TRANSACTIONS ON COMPONENTS PACKAGING AND MANUFACTURING TECHNOLOGY
(2022)
Article
Computer Science, Interdisciplinary Applications
Bence Tipary, Gabor Erdos
Summary: This paper proposes a generalized development methodology for flexible robotic pick-and-place workcells based on the concept of Digital Twin, aiming to speed up the overall commissioning process and reduce the amount of work in the physical workcell.
ROBOTICS AND COMPUTER-INTEGRATED MANUFACTURING
(2021)
Article
Robotics
Fengyi Wang, Julio Rogelio Guadarrama-Olvera, Gordon Cheng
Summary: Our fast method for autonomously planning manipulation tasks for mobile manipulators effectively minimizes execution time and is suitable for various scenarios.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2021)
Article
Robotics
Hao Chen, Takuya Kiyokawa, Weiwei Wan, Kensuke Harada
Summary: This article introduces a new method for robotic pick-and-place tasks, achieving high accuracy and stabilization through similarity matching and category association. By using word embedding to quantify the similarity and planning grasps on real-world objects, the limitations of learning-based approaches are effectively addressed.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2022)
Article
Robotics
Marcus Gualtieri, Robert Platt
Summary: The study focuses on robotic pick-and-place of partially visible, novel objects and proposes a method for handling uncertainty in object shapes, incorporating it into the cost function. Among seven different costs compared, using neural networks to estimate grasp and place stability probability consistently outperforms others, achieving a success rate of 7.8% higher than the commonly used minimum-number-of-grasps cost when packing objects tightly into a bin with a real robot.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2021)
Article
Engineering, Multidisciplinary
Guoyu Zuo, Mi Li, Jianjun Yu, Chun Wu, Gao Huang
Summary: This paper proposes a novel motion planner called the Lazy Demonstration Graph (LDG) planner, which exploits successful planning cases as prior knowledge and adapts to workplace changes through adaptive sampling strategy and lazy collision detection. Compared with other motion planning algorithms, our method achieves better performance in planning time and path quality.
Article
Robotics
Francesco Di Felice, Salvatore D'Avella, Alberto Remus, Paolo Tripicchio, Carlo Alberto Avizzano
Summary: This paper proposes a framework based on Graph Neural Networks (GNN) that allows a robot to learn task-specific rules from synthetic demonstrations through imitation learning. The task is abstracted and represented as a graph, with task-relevant entities encoded as nodes. The GNN-based policy learns the underlying rules of the manipulation task, focusing on structural relevance and object type, while relying on an external primitive for robot movement. Experimental results show that the proposed model has high generalization capability, accommodating various object/goal configurations.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2023)
Article
Robotics
Tingting Su, Xu Liang, Xiang Zeng, Shengda Liu
Summary: This paper proposes a Pythagorean-Hodograph (PH) curve-based pick-and-place operation trajectory planning method for Delta parallel robots, which achieves flexible control of pick-and-place operations to meet various practical scenarios. Different pick-and-place operations are classified based on the geometric relationship of the operation path. Trajectory planning is then carried out for each situation, with different polynomial motion laws solved for the linear and curved motion segments. Trajectory optimization is performed with the motion period as the objective. The proposed method realizes obstacle avoidance, optimal time, flexible control of robot trajectory, and stable motion.
Article
Energy & Fuels
Lukasz Gruszka, Michal Bartys
Summary: This paper aims to present a rational and systematic approach to reduce energy consumption in pick-and-place operations performed by robots. By describing a new path planning method, it achieves the minimization of energy consumption while preserving productivity and taking care of the mechanical components' persistence in the robot cooperating with the autonomous mobile platform.
Article
Computer Science, Artificial Intelligence
Zhengkai Li, Hao Sun, Xinghu Yu, Weichao Sun
Summary: This paper investigates the capacitated location routing problems (CLRP) and how to solve them using an improved Hopfield neural network (HNN). The original problem is decomposed into three subproblems and solved using a heuristic sequencing HNN approach. Experimental results demonstrate the effectiveness and efficiency of this method on practical industrial data.