Review
Engineering, Manufacturing
Thomas Feldhausen, Lauren Heinrich, Kyle Saleeby, Alan Burl, Brian Post, Eric MacDonald, Chris Saldana, Lonnie Love
Summary: Hybrid additive manufacturing combines the advantages of additive and subtractive manufacturing to produce parts with complex geometries, improved surface finish, and precise dimensional accuracy. CAM software plays a crucial role in orchestrating the machine toolpathing for both deposition and machining processes. Different hybrid systems can benefit from optimized toolpath planning to improve the microstructure, mechanical properties, porosity, and residual stress of the final fabricated part.
ADDITIVE MANUFACTURING
(2022)
Article
Automation & Control Systems
Elie Maalouf, Joanna Daaboul, Julien Le Duigou, Bassam Hussein
Summary: This paper introduces a distributed approach for smart production management in a cellular manufacturing system, with decision levels at the factory, shop floor, and cell levels. The approach integrates planning, scheduling, and material handling allocation while considering real-time data from the supply chain. Testing and optimization were conducted using a mathematical model and two metaheuristics.
INTERNATIONAL JOURNAL OF ADVANCED MANUFACTURING TECHNOLOGY
(2022)
Article
Health Care Sciences & Services
Paolo Scolozzi, Francesco Michelini, Claude Crottaz, Alexandre Perez
Summary: Computer-aided design and computer-aided manufacturing (CAD/CAM) techniques have proven to be valuable in dental implant surgery, improving the success rate and patient satisfaction while ensuring implant parallelism. This study analyzed the clinical and radiological data of thirteen edentulous patients treated using CAD/CAM techniques, showing successful outcomes.
JOURNAL OF PERSONALIZED MEDICINE
(2023)
Article
Engineering, Chemical
Angela Luft, Sebastian Bremen, Nils Luft
Summary: There is a growing demand for flexibility in manufacturing to adapt to market volatility and provide customization for customers. Additive manufacturing (AM) is being considered as a solution to make production more flexible. This paper develops a conceptual model and application guideline to quantify the impact of AM on volume and mix flexibility in production systems. A case study is presented to demonstrate the potential impact of additive technologies on manufacturing flexibility. This work allows factory planners to assess the effects of AM on flexibility and design their production systems accordingly.
Article
Engineering, Industrial
Quentin Christ, Stephane Dauzere-Peres, Guillaume Lepelletier
Summary: This paper addresses a practical operational production planning problem in complex manufacturing systems and introduces a three-step approach to solve the problem. The approach is further optimized by introducing new smoothing rules and their performance is studied. The paper also presents how the decision support tool is embedded in the production planning process to bridge the gap between the upper and lower planning levels.
INTERNATIONAL JOURNAL OF PRODUCTION RESEARCH
(2023)
Article
Chemistry, Multidisciplinary
Kyu Tae Park, Yoo Ho Son, Sang Wook Ko, Sang Do Noh
Summary: To achieve efficient personalized production at an affordable cost, modular manufacturing systems (MMS) can be used, with a micro smart factory (MSF) being an example of MMS with heterogeneous production processes. However, MSFs need to overcome performance hurdles with respect to production control, which is why a digital twin (DT) and reinforcement learning (RL)-based production control method is proposed in this paper to provide a resilient solution for the cyber-physical production systems (CPPSs) architectural framework and to make appropriate decisions in dynamic production situations.
APPLIED SCIENCES-BASEL
(2021)
Article
Chemistry, Analytical
Yupeng Xin, Yiwen Chen, Wenhui Li, Xiuhong Li, Fengfeng Wu
Summary: This article introduces a method of using digital twin technology to build a high-fidelity process model and optimize process plans. By feeding surface inspection data of parts back to the CAPP system and associating it with the digital twin process model, and then using the Poisson reconstruction algorithm for simulation, it is found that the digital twin model can improve simulation accuracy.
Article
Engineering, Industrial
Milad Elyasi, Basak Altan, Ali Ekici, Okan Orsan Ozener, Ihsan Yanikoglu, Alexandre Dolgui
Summary: This paper examines the impact of the global crisis on supply chain resilience and suggests the implementation of flexible/hybrid manufacturing systems as a viable strategy. Using Vestel Electronics as a case study, the research proposes a flexible/hybrid manufacturing production setup to address uncertain demand. By employing a scenario-based approach and a heuristic algorithm based on column generation, the optimization model demonstrates effective and cost-efficient solutions.
INTERNATIONAL JOURNAL OF PRODUCTION RESEARCH
(2023)
Article
Computer Science, Interdisciplinary Applications
Hitesh Dhiman, Carsten Roecker
Summary: This paper introduces MiWSICx, a middleware that combines multiple interactive computing devices to provide crossdevice industrial assistance. Based on activity theory, it models human work activities combining multiple users, artifacts, and cyber-physical objects, using the actor model for deployment on various hardware to provide multiuser, crossdevice, multiactivity assistance within CPPSs.
JOURNAL OF COMPUTATIONAL DESIGN AND ENGINEERING
(2021)
Article
Engineering, Industrial
Hubert Missbauer, Reha Uzsoy
Summary: Production planning and control systems are crucial for the competitiveness of manufacturing firms and their ability to leverage technological advancements. The research on order release function within PPC systems reveals a need for optimization models and further understanding of the dynamic response of production units. There are still unresolved research questions in this area that warrant further exploration.
INTERNATIONAL JOURNAL OF PRODUCTION RESEARCH
(2022)
Article
Computer Science, Interdisciplinary Applications
Veronica Lindstrom, Fredrik Persson, Arun Pravin Chennai Viswanathan, Mahendran Rajendran
Summary: This paper explores the relationship between the four elements of smart production planning and control (PPC) and data quality issues, and provides a framework to address these problems. The empirical data results show that inaccurate data entries are the most common data quality problem in PPC, with causes linked to human resources and organizational control. Maintaining a high level of data quality is crucial for achieving high-performance production planning and control.
COMPUTERS IN INDUSTRY
(2023)
Article
Computer Science, Interdisciplinary Applications
Aman Kukreja, S. S. Pande
Summary: This paper presents a novel voxel-based toolpath planning algorithm for three-axis milling of freeform surfaces, aiming to address the computational complexity and time consumption issues. Two strategies, iso-scallop and hybrid iso-scallop, were proposed and tested extensively for complex freeform surface parts. The developed system was found to be computationally efficient, robust, and accurate in generating a finishing toolpath.
JOURNAL OF COMPUTING AND INFORMATION SCIENCE IN ENGINEERING
(2023)
Article
Automation & Control Systems
Emre Yildiz, Charles Moller, Arne Bilberg
Summary: Smart manufacturing, driven by the 4th industrial revolution and forces like innovation, competition, and changing demands, requires manufacturing companies to reform and regenerate their product, process, and system models to stay competitive. The digital twin-based virtual factory concept shows potential in supporting manufacturing organizations to adapt to dynamic and complex environments through virtual collaboration and prototyping.
INTERNATIONAL JOURNAL OF ADVANCED MANUFACTURING TECHNOLOGY
(2021)
Article
Computer Science, Artificial Intelligence
Amon Goeppert, Lea Grahn, Jonas Rachner, Dennis Grunert, Simon Hort, Robert H. Schmitt
Summary: The demand for individualized products drives the development of manufacturing systems towards greater adaptability and flexibility, with digital twins serving as a fully connected digital model to physical and digital assets. Standardization and structured modeling are crucial in the creation and deployment of digital twins, along with communication standards and protocols for data exchange. However, there is a lack of consistent workflow from ontology-driven definition to standardized modeling. This paper aims to design an end-to-end digital twin pipeline and automate the process of establishing communication connections. A line-less assembly system with manual stations and a mobile robot is used as an example to explain the digital twin pipeline transparently.
JOURNAL OF INTELLIGENT MANUFACTURING
(2023)
Review
Engineering, Industrial
Jacob Lohmer, Rainer Lasch
Summary: Recent research on multi-factory production planning and scheduling problems has focused on addressing market uncertainty and technological trends. Scholars are exploring ways to efficiently manage geographically dispersed factories and allocate orders and tasks effectively.
INTERNATIONAL JOURNAL OF PRODUCTION RESEARCH
(2021)
Article
Computer Science, Artificial Intelligence
Mahmoud Akl, Deniz Ergene, Florian Walter, Alois Knoll
Summary: Deep reinforcement learning (DRL) combines reinforcement learning algorithms with deep neural networks (DNNs). Spiking neural networks (SNNs) have been shown to be a biologically plausible and energy efficient alternative to DNNs. They can be trained with the backpropagation through time (BPTT) algorithm. SNNs are sensitive to hyperparameters introduced by spiking neuron models, but increasing simulation time and applying a two-neuron encoding to the input observations can reduce this sensitivity.
FRONTIERS IN NEUROROBOTICS
(2023)
Article
Automation & Control Systems
Zhenshan Bing, Hongkuan Zhou, Rui Li, Xiaojie Su, Fabrice O. Morin, Kai Huang, Alois Knoll
Summary: This article introduces a method called GC-HGG, an extension of hindsight goal generation, which selects hindsight goals based on graph-based proximity and diversity. It was evaluated in four challenging manipulation tasks and demonstrated significant improvements in both sample efficiency and overall success rates.
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS
(2023)
Article
Engineering, Civil
Xinyi Li, Yinlong Liu, Venkatnarayanan Lakshminarasimhan, Hu Cao, Feihu Zhang, Alois Knoll
Summary: In this paper, a targetless traffic radar calibration method based on GPS is proposed to overcome the inconvenience during ITS operation. A high-precision GPS device installed on the moving vehicle provides accurate positioning information of the detection target. A globally optimal registration method, called Gaussian Mixture Robust Branch and Bound (GMRBnB), is proposed for the optimization process of extrinsic calibration, which is robust to noise and outliers in radar measurements.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2023)
Article
Computer Science, Theory & Methods
Patrik Omland, Alessio Netti, Yang Peng, Andrea Baldovin, Michael Paulitsch, Gustavo Espinosa, Jorge Parra, Gereon Hinz, Alois Knoll
Summary: Chips are getting denser with increasingly smaller transistors, leading to higher fault rates, especially in High-Performance Computing (HPC) systems. While hardware fault protection is expensive and HPC users expect correct application output, tolerances for output errors are allowed. To address this, a user-centric reliability benchmark is proposed to specify HPC system reliability targets, allowing for better performance optimizations in hardware design. The open-source Hardware Design Fault Injection Toolkit (HDFIT) enables end-to-end hardware design reliability experiments, including netlist-level fault injection and application output error analysis. A proof-of-concept study on the reliability of an open-source GEMM accelerator targeting popular applications is presented using HDFIT.
IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS
(2023)
Article
Engineering, Electrical & Electronic
Jiayi Guan, Guang Chen, Jin Huang, Zhijun Li, Lu Xiong, Jing Hou, Alois Knoll
Summary: Autonomous driving is a promising technology for reducing traffic accidents and improving driving efficiency. In this study, we propose a discrete decision-making strategy based on the DSAC-SF algorithm to enhance driving efficiency and safety on freeways. Experimental results demonstrate that our strategy achieves a high success rate and fast vehicle speed in decision-making tasks, while our DSAC-SF algorithm shows improved training efficiency and stability compared to commonly used discrete reinforcement learning algorithms.
IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY
(2023)
Article
Automation & Control Systems
Hu Cao, Guang Chen, Zhijun Li, Qian Feng, Jianjie Lin, Alois Knoll
Summary: This study proposes an efficient grasp detection network for robotic grasp tasks. The network uses a lightweight generative structure to achieve a balance between high grasp confidence and fast inference speed. It introduces a Gaussian kernel-based grasp representation for encoding training samples and employs receptive field blocks and attention mechanisms for improved feature discriminability and semantic information fusion. Experimental results demonstrate excellent performance on Cornell, Jacquard, and extended OCID grasp datasets with accuracy of 97.8%, 95.6%, and 76.4% respectively.
IEEE-ASME TRANSACTIONS ON MECHATRONICS
(2023)
Editorial Material
Computer Science, Artificial Intelligence
Zhenshan Bing, Chenguang Yang, Alois Knoll
FRONTIERS IN NEUROROBOTICS
(2023)
Article
Computer Science, Artificial Intelligence
Yusuke Kuniyoshi, Rin Kuriyama, Shu Omura, Carlos Enrique Gutierrez, Zhe Sun, Benedikt Feldotto, Ugo Albanese, Alois C. Knoll, Taiki Yamada, Tomoya Hirayama, Fabrice O. Morin, Jun Igarashi, Kenji Doya, Tadashi Yamazaki
Summary: This study presents the development of a spiking brain model and a mouse body model, connected over the Internet and run in a distributed manner, demonstrating the potential and usefulness of distributed embodied simulation in understanding animal behavior and behavioral change.
FRONTIERS IN NEUROROBOTICS
(2023)
Article
Robotics
Yuning Cui, Alois Knoll
Summary: Image restoration is crucial in various fields such as robot vision, autonomous vehicles, and medical imaging, to recover clear images from degradations. The use of Transformer has shown significant improvement, but its practicality is limited due to high complexity. To address this, we propose an efficient image restoration framework based on self-attention, combining patch-based and strip-based units for improved efficiency. Our framework, PSNet, achieves state-of-the-art performance on multiple restoration tasks with low computational complexity and high speed.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2023)
Proceedings Paper
Computer Science, Artificial Intelligence
Long Wen, Markus Rickert, Fengjunjie Pan, Jianjie Lin, Alois Knoll
Summary: Software-defined vehicles (SDV) are crucial for the rapid development of autonomous driving due to their increased flexibility compared to traditional architectures. Containerization and virtualization are key technologies enabling fast software installation and updates under the SDV framework. However, their performance and suitability for intelligent vehicles still need to be evaluated.
2023 IEEE INTELLIGENT VEHICLES SYMPOSIUM, IV
(2023)
Article
Computer Science, Artificial Intelligence
Mingchuan Zhou, Xiangyu Guo, Matthias Grimm, Elias Lochner, Zhongliang Jiang, Abouzar Eslami, Juan Ye, Nassir Navab, Alois Knoll, Mohammad Ali Nasseri
Summary: In this paper, a robust framework is proposed for needle detection and localisation in robot-assisted subretinal injection using microscope-integrated Optical Coherence Tomography with deep learning. Five different architectures of convolutional neural networks were evaluated, and the top performing network successfully detected all needles and localised them with an Intersection over Union value of 0.55. The algorithm was evaluated by comparing the depth of the predicted bounding box to the ground truth, and the top edge was found to have a maximum error of 8.5 mu m in predicting the depth of the needle.
CAAI TRANSACTIONS ON INTELLIGENCE TECHNOLOGY
(2023)
Article
Computer Science, Artificial Intelligence
Zhenshan Bing, David Lerch, Kai Huang, Alois Knoll
Summary: The subject of deep reinforcement learning (DRL) has developed rapidly and is now applied in various fields. However, artificial agents trained with RL algorithms require large amounts of training data. The concept of meta-reinforcement learning (meta-RL) enables agents to learn new skills from a small amount of experience. This study introduces a training strategy for non-stationary environments and a task representation based on Gaussian mixture models.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2023)
Proceedings Paper
Computer Science, Artificial Intelligence
Etienne Mueller, Daniel Auge, Alois Knoll
Summary: As machine learning applications continue to grow, the need for energy-efficient implementations is rising. A promising approach is the use of spiking neural networks with neuromorphic hardware, as energy is only consumed during information processing. However, there are challenges in creating a uniform threshold for spiking neurons. Previous work mainly focused on feedforward or convolutional networks, making it difficult to process continuous sequential data. This study presents a novel method to convert a long short-term memory network to a spiking neural network, leveraging a population of spiking neurons with a normally distributed threshold to represent activation functions.
ADVANCES IN NATURAL COMPUTATION, FUZZY SYSTEMS AND KNOWLEDGE DISCOVERY, ICNC-FSKD 2022
(2023)
Article
Chemistry, Analytical
Jianguo Hu, Goektug Yeslbs, Yuanyuan Li, Xiangshun Geng, Jing Chen, Xiaoming Wu, Alois Knoll, Tian-Ling Ren
Summary: In this study, the overtone mass sensitivity of a QCM sensor with an asymmetric N-M electrode configuration was measured through electrochemical electrodeposition experiments. The results showed that the overtone mass sensitivity of the N-M type QCM is larger than that of QCMs with symmetric electrodes, and the fifth overtone mass sensitivity is higher than the third overtone mass sensitivity for the same type of QCM.
ANALYTICAL CHEMISTRY
(2023)
Article
Computer Science, Information Systems
Tapisha Soni, Malte Schellmann, Alois C. Knoll
Summary: In this paper, a new scheme based on 5G networks is proposed to improve communication reliability in future factories. By enhancing D2D connections and utilizing SL-assisted retransmissions, the proposed scheme achieves significant reduction in retransmission count and power consumption, as well as resource reuse. System-level simulations in a realistic factory scenario validate the effectiveness of the scheme.