Procedia CIRP最新文献_第10页

A prediction method of surface geometric deviation for additive manufacturing parts based on knowledge-integrated deep learning algorithm 基于知识集成深度学习算法的增材制造零件表面几何偏差预测方法

Procedia CIRP

Pub Date : 2024-01-01 DOI: 10.1016/j.procir.2024.10.005

Zhicheng Huang , Yingyu Cao , Yuda Cao , Kai Guo , Lihong Qiao

Compared with traditional machining processes, additive manufacturing (AM) has received widespread attention in recent years because of its high degree of modeling freedom. However, due to the multiple manufacturing errors and complex physical state changes involved in the process, the geometric deviation on the AM part surface is a challenge for controlling product geometrical quality. To address this problem, data-driven machine learning (ML) techniques have been widely studied in product quality controlling. However, traditional ML greatly depends on the training sample data, and suffers the risk of violating physical mechanisms due to the lack of domain knowledge. In order to take the best advantage of domain knowledge, prior information and deep learning algorithm, this paper proposes a knowledge-integrated deep learning algorithm and constructs the geometric deviation prediction model of the AM part surface. After that, the method was verified with design of experiments. The results show that compared with the data-driven neural network (DDNN), the knowledge-integrated neural network (KINN) has fewer iterations during the training process, less sample data requirement and more accurate prediction results.

与传统加工工艺相比，快速成型制造（AM）因其建模自由度高，近年来受到广泛关注。然而，由于加工过程中存在多种制造误差和复杂的物理状态变化，AM 零件表面的几何偏差成为控制产品几何质量的难题。为解决这一问题，数据驱动的机器学习（ML）技术已在产品质量控制领域得到广泛研究。然而，传统的 ML 在很大程度上依赖于训练样本数据，并且由于缺乏领域知识而存在违反物理机制的风险。为了充分利用领域知识、先验信息和深度学习算法，本文提出了一种知识集成深度学习算法，并构建了 AM 零件表面几何偏差预测模型。随后，通过实验设计对该方法进行了验证。结果表明，与数据驱动神经网络（DDNN）相比，知识集成神经网络（KINN）在训练过程中的迭代次数更少，样本数据要求更低，预测结果更准确。

{"title":"A prediction method of surface geometric deviation for additive manufacturing parts based on knowledge-integrated deep learning algorithm","authors":"Zhicheng Huang , Yingyu Cao , Yuda Cao , Kai Guo , Lihong Qiao","doi":"10.1016/j.procir.2024.10.005","DOIUrl":"10.1016/j.procir.2024.10.005","url":null,"abstract":"<div><div>Compared with traditional machining processes, additive manufacturing (AM) has received widespread attention in recent years because of its high degree of modeling freedom. However, due to the multiple manufacturing errors and complex physical state changes involved in the process, the geometric deviation on the AM part surface is a challenge for controlling product geometrical quality. To address this problem, data-driven machine learning (ML) techniques have been widely studied in product quality controlling. However, traditional ML greatly depends on the training sample data, and suffers the risk of violating physical mechanisms due to the lack of domain knowledge. In order to take the best advantage of domain knowledge, prior information and deep learning algorithm, this paper proposes a knowledge-integrated deep learning algorithm and constructs the geometric deviation prediction model of the AM part surface. After that, the method was verified with design of experiments. The results show that compared with the data-driven neural network (DDNN), the knowledge-integrated neural network (KINN) has fewer iterations during the training process, less sample data requirement and more accurate prediction results.</div></div>","PeriodicalId":20535,"journal":{"name":"Procedia CIRP","volume":"129 ","pages":"Pages 19-24"},"PeriodicalIF":0.0,"publicationDate":"2024-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142555682","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Price competition for service provision with different bargaining abilities 不同议价能力的服务价格竞争

Procedia CIRP

Pub Date : 2024-01-01 DOI: 10.1016/j.procir.2024.08.253

Jianyong Teng , Ryuichiro Ishikawa

Price negotiation still has an important role in business-to-business trades. In the negotiations, the realized price depends on the bargaining power of each firm. Therefore, it is important for consumers to know which firm has weaker bargaining power. Simultaneously, some consumers do not negotiate at all. In the case, they buy goods at the price that firms publicly post. In the sense, two different pricing system, the negotiated price and posted price, consumers exist in trades. In our paper, we also consider that the consumers need to pay their transportation costs. In this setting, we analyze how consumers determine which firm each of them visit, and how the realized price is formed. Our results show that negotiated prices depends on their opponent bargaining powers, but independent from consumers’ transportation costs.

价格谈判在企业对企业的贸易中仍然发挥着重要作用。在谈判中，实现的价格取决于每家公司的议价能力。因此，消费者必须知道哪家公司的议价能力较弱。同时，有些消费者根本不谈判。在这种情况下，他们按照企业公开发布的价格购买商品。从这个意义上说，消费者在交易中存在两种不同的定价体系，即协商价格和公布价格。在本文中，我们还考虑到消费者需要支付运输成本。在这种情况下，我们分析了消费者如何决定各自访问哪家公司，以及实现价格是如何形成的。我们的结果表明，协商价格取决于双方的讨价还价能力，但与消费者的运输成本无关。

引用次数: 0

Digital Battery Cell Manufacturing Systems: Approaching a scalable and versatile IT Architecture Design 数字电池芯制造系统：接近可扩展的多功能 IT 架构设计

Procedia CIRP

Pub Date : 2024-01-01 DOI: 10.1016/j.procir.2024.10.119

Leon Mohring , Wilhelm Jaspers , Arno Schmetz , David Roth , Nils Christian Hamacher , Achim Kampker

The transition from a fossil-fuel powered economy towards decentralized renewable energy sources and electric mobility creates a global demand for battery cells. As cell manufacturers ramp up production capacity, they are facing the challenge of scaling their IT systems in pace with the growing demand. Innovations such as new cell types and enhanced production technology as well as incorporating ever-evolving data-driven methods, e.g., Artificial intelligence, require a versatile IT architecture, capable of adapting. In the light of these challenges, this paper introduces a methodology aimed at building an IT architecture tailored to the requirements of a given battery cell manufacturing use case. It provides different approaches with respect to several dimensions, considering among others the requirements of shopfloor connectivity, data acquisition and storage strategies, as well as IT systems integration design. Further covered requirements deal with computational capacities close to the shopfloor, real-time aspects, data ingestion and integration, and dynamic resource allocation. This paper presents the application of the methodology to a multi-site high-scale production at Fraunhofer FFB, describing the resulting IT architecture. The factories combined will host four manufacturing lines producing a GWh-scale battery cell output per year. To sustain the workload on the FFB IT architecture incurred by this throughput, the design of the IT architecture pivots away from a strictly hierarchical structure, bringing critical systems and databases closer to the shopfloor. This and other design choices have been made based on the developed methodology. Overall, the presented methodology and the derived FFB IT architecture show a path towards building a battery cell production IT architecture capable of scaling and rapidly adapting to new technologies.

从化石燃料驱动的经济向分散式可再生能源和电动汽车的过渡，创造了对电池的全球需求。随着电池制造商提高产能，他们正面临着如何扩展其 IT 系统以满足日益增长的需求的挑战。新型电池和增强型生产技术等创新，以及不断发展的数据驱动方法（如人工智能），都需要一个能够适应的多功能 IT 架构。考虑到这些挑战，本文介绍了一种方法，旨在根据特定电池制造用例的要求建立一个 IT 架构。它从多个方面提供了不同的方法，主要考虑了车间连接、数据采集和存储策略以及 IT 系统集成设计等方面的要求。此外，该方法还涉及车间附近的计算能力、实时性、数据采集和集成以及动态资源分配等方面的要求。本文介绍了该方法在弗劳恩霍夫 FFB 公司多站点大规模生产中的应用，并描述了由此产生的 IT 架构。这些工厂将拥有四条生产线，每年生产 GWh 规模的电池片。为了承受这种产量给 FFB IT 架构带来的工作量，IT 架构的设计不再采用严格的分级结构，而是将关键系统和数据库更靠近车间。这一设计选择和其他设计选择都是根据所开发的方法做出的。总之，所介绍的方法和衍生的 FFB IT 架构为建立一个能够扩展和快速适应新技术的电池生产 IT 架构指明了道路。

{"title":"Digital Battery Cell Manufacturing Systems: Approaching a scalable and versatile IT Architecture Design","authors":"Leon Mohring , Wilhelm Jaspers , Arno Schmetz , David Roth , Nils Christian Hamacher , Achim Kampker","doi":"10.1016/j.procir.2024.10.119","DOIUrl":"10.1016/j.procir.2024.10.119","url":null,"abstract":"<div><div>The transition from a fossil-fuel powered economy towards decentralized renewable energy sources and electric mobility creates a global demand for battery cells. As cell manufacturers ramp up production capacity, they are facing the challenge of scaling their IT systems in pace with the growing demand. Innovations such as new cell types and enhanced production technology as well as incorporating ever-evolving data-driven methods, e.g., Artificial intelligence, require a versatile IT architecture, capable of adapting. In the light of these challenges, this paper introduces a methodology aimed at building an IT architecture tailored to the requirements of a given battery cell manufacturing use case. It provides different approaches with respect to several dimensions, considering among others the requirements of shopfloor connectivity, data acquisition and storage strategies, as well as IT systems integration design. Further covered requirements deal with computational capacities close to the shopfloor, real-time aspects, data ingestion and integration, and dynamic resource allocation. This paper presents the application of the methodology to a multi-site high-scale production at Fraunhofer FFB, describing the resulting IT architecture. The factories combined will host four manufacturing lines producing a GWh-scale battery cell output per year. To sustain the workload on the FFB IT architecture incurred by this throughput, the design of the IT architecture pivots away from a strictly hierarchical structure, bringing critical systems and databases closer to the shopfloor. This and other design choices have been made based on the developed methodology. Overall, the presented methodology and the derived FFB IT architecture show a path towards building a battery cell production IT architecture capable of scaling and rapidly adapting to new technologies.</div></div>","PeriodicalId":20535,"journal":{"name":"Procedia CIRP","volume":"130 ","pages":"Pages 492-497"},"PeriodicalIF":0.0,"publicationDate":"2024-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142723406","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

The role of maintenance in company-specific production systems 维护在公司特定生产系统中的作用

Procedia CIRP

Pub Date : 2024-01-01 DOI: 10.1016/j.procir.2024.10.133

Anderson Leal , Jon Bokrantz , Anders Skoogh

Company-Specific Production Systems (XPS) are in use in several manufacturing companies around the globe, as a reference to variations of the Toyota Production System. XPS is a continuous improvement program responsible for increasing the general performance of companies. Maintenance has an important contribution to XPS by delivering technical availability at a rational cost. However, the connections between all the elements of the XPS and the corresponding contributions from maintenance are not crystal clear. Providing such clarity could increase the focus on improvements that would create real benefits for the company. The current study aims to bridge the XPS literature to maintenance applications, thereby substantiating the role of maintenance in XPS. Firstly, a theoretical framework of XPS is created and explained based on previous literature. The framework outlines three core elements of an XPS: content, management, and outcomes. Also, it presents the interconnections between the elements. Secondly, the framework acts as a guide to an empirical study at an automotive company in Sweden. The study maps the role of maintenance and its contribution to the XPS in place. For each of the XPS elements, a maintenance correspondent was selected and connected to the XPS framework. Thirdly, based on the results of the empirical study, the paper proposes a set of critical research directions, both guiding the design and execution of future research studies and supporting the long-term competitiveness of the company.

公司特定生产系统（XPS）是丰田生产系统的变体，目前已在全球多家制造公司使用。XPS 是一项持续改进计划，负责提高企业的总体绩效。通过以合理的成本提供技术可用性，维护对 XPS 有着重要的贡献。然而，XPS 的所有要素与维护的相应贡献之间的联系并不十分清晰。提供这样的清晰度可以使人们更加关注能为公司创造实际效益的改进措施。本研究旨在将 XPS 文献与维护应用联系起来，从而证实维护在 XPS 中的作用。首先，在以往文献的基础上建立并解释了 XPS 的理论框架。该框架概述了 XPS 的三个核心要素：内容、管理和结果。此外，它还介绍了各要素之间的相互联系。其次，该框架为瑞典一家汽车公司的实证研究提供了指导。该研究描绘了维护的作用及其对现有 XPS 的贡献。针对每个 XPS 要素，都选择了一个维护对应方，并将其与 XPS 框架连接起来。第三，根据实证研究的结果，本文提出了一系列重要的研究方向，既能指导未来研究的设计和实施，又能为公司的长期竞争力提供支持。

{"title":"The role of maintenance in company-specific production systems","authors":"Anderson Leal , Jon Bokrantz , Anders Skoogh","doi":"10.1016/j.procir.2024.10.133","DOIUrl":"10.1016/j.procir.2024.10.133","url":null,"abstract":"<div><div>Company-Specific Production Systems (XPS) are in use in several manufacturing companies around the globe, as a reference to variations of the Toyota Production System. XPS is a continuous improvement program responsible for increasing the general performance of companies. Maintenance has an important contribution to XPS by delivering technical availability at a rational cost. However, the connections between all the elements of the XPS and the corresponding contributions from maintenance are not crystal clear. Providing such clarity could increase the focus on improvements that would create real benefits for the company. The current study aims to bridge the XPS literature to maintenance applications, thereby substantiating the role of maintenance in XPS. Firstly, a theoretical framework of XPS is created and explained based on previous literature. The framework outlines three core elements of an XPS: content, management, and outcomes. Also, it presents the interconnections between the elements. Secondly, the framework acts as a guide to an empirical study at an automotive company in Sweden. The study maps the role of maintenance and its contribution to the XPS in place. For each of the XPS elements, a maintenance correspondent was selected and connected to the XPS framework. Thirdly, based on the results of the empirical study, the paper proposes a set of critical research directions, both guiding the design and execution of future research studies and supporting the long-term competitiveness of the company.</div></div>","PeriodicalId":20535,"journal":{"name":"Procedia CIRP","volume":"130 ","pages":"Pages 584-590"},"PeriodicalIF":0.0,"publicationDate":"2024-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142723422","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Concept drift monitoring for industrial load forecasting with artificial neural networks 利用人工神经网络进行工业负荷预测的概念漂移监测

Procedia CIRP

Pub Date : 2024-01-01 DOI: 10.1016/j.procir.2024.10.065

Robin Zink , Borys Ioshchikhes , Matthias Weigold

Long Short-Term Memory (LSTM) models are frequently applied for industrial energy load forecasting. However, real-world production systems are highly dynamic which poses major challenges. Concept drifts potentially lead to performance degradation that affects systems optimization for the worse. In this work, Concept Drift Detection (CDD) for industrial energy load forecasting with LSTM models is researched. For this purpose, five CDD algorithms are evaluated using the active power of a machine tool. Drift Detection Method (DDM) and Kolmogorov-Smirnov Windowing (KSWIN) proved to be particularly effective with easily interpretable and reasonable hyperparameters.

长短期记忆（LSTM）模型经常被用于工业能源负荷预测。然而，现实世界的生产系统是高度动态的，这就带来了重大挑战。概念漂移可能会导致性能下降，从而对系统的优化产生负面影响。在这项工作中，研究了利用 LSTM 模型进行工业能源负荷预测的概念漂移检测（CDD）。为此，利用机床的有功功率对五种 CDD 算法进行了评估。事实证明，漂移检测法（DDM）和 Kolmogorov-Smirnov Windowing (KSWIN) 特别有效，其超参数易于解释且合理。

引用次数: 0

Industrial Language-Image Dataset (ILID): Adapting Vision Foundation Models for Industrial Settings 工业语言图像数据集 (ILID)：针对工业环境调整视觉基础模型

Procedia CIRP

Pub Date : 2024-01-01 DOI: 10.1016/j.procir.2024.10.084

Keno Moenck , Duc Trung Thieu , Julian Koch , Thorsten Schüppstuhl

In recent years, the upstream of Large Language Models (LLM) has also encouraged the computer vision community to work on substantial multimodal datasets and train models on a scale in a self-/semi-supervised manner, resulting in Vision Foundation Models (VFM), as, e.g., Contrastive Language–Image Pre-training (CLIP). The models generalize well and perform outstandingly on everyday objects or scenes, even on downstream tasks, tasks the model has not been trained on, while the application in specialized domains, as in an industrial context, is still an open research question. Here, fine-tuning the models or transfer learning on domain-specific data is unavoidable when objecting to adequate performance. In this work, we, on the one hand, introduce a pipeline to generate the Industrial Language-Image Dataset (ILID) based on web-crawled data; on the other hand, we demonstrate effective self-supervised transfer learning and discussing downstream tasks after training on the cheaply acquired ILID, which does not necessitate human labeling or intervention. With the proposed approach, we contribute by transferring approaches from state-of-the-art research around foundation models, transfer learning strategies, and applications to the industrial domain.

近年来，大型语言模型（LLM）的发展也鼓励了计算机视觉界研究大量的多模态数据集，并以自我/半监督的方式对模型进行大规模训练，从而产生了视觉基础模型（VFM），例如对比语言-图像预训练（CLIP）。这些模型具有良好的通用性，在日常物体或场景中表现出色，甚至在下游任务、模型未经过训练的任务中也是如此，而在专业领域（如工业环境）中的应用仍是一个有待解决的研究问题。在这种情况下，要想获得足够的性能，对模型进行微调或对特定领域的数据进行迁移学习是不可避免的。在这项工作中，我们一方面介绍了一种基于网络抓取数据生成工业语言图像数据集（ILID）的管道；另一方面，我们展示了有效的自监督迁移学习，并讨论了在廉价获取的 ILID 上进行训练后的下游任务，这种训练无需人工标注或干预。通过所提出的方法，我们将围绕基础模型、迁移学习策略和应用的最新研究方法迁移到了工业领域。

{"title":"Industrial Language-Image Dataset (ILID): Adapting Vision Foundation Models for Industrial Settings","authors":"Keno Moenck , Duc Trung Thieu , Julian Koch , Thorsten Schüppstuhl","doi":"10.1016/j.procir.2024.10.084","DOIUrl":"10.1016/j.procir.2024.10.084","url":null,"abstract":"<div><div>In recent years, the upstream of Large Language Models (LLM) has also encouraged the computer vision community to work on substantial multimodal datasets and train models on a scale in a self-/semi-supervised manner, resulting in Vision Foundation Models (VFM), as, e.g., Contrastive Language–Image Pre-training (CLIP). The models generalize well and perform outstandingly on everyday objects or scenes, even on downstream tasks, tasks the model has not been trained on, while the application in specialized domains, as in an industrial context, is still an open research question. Here, fine-tuning the models or transfer learning on domain-specific data is unavoidable when objecting to adequate performance. In this work, we, on the one hand, introduce a pipeline to generate the Industrial Language-Image Dataset (ILID) based on web-crawled data; on the other hand, we demonstrate effective self-supervised transfer learning and discussing downstream tasks after training on the cheaply acquired ILID, which does not necessitate human labeling or intervention. With the proposed approach, we contribute by transferring approaches from state-of-the-art research around foundation models, transfer learning strategies, and applications to the industrial domain.</div></div>","PeriodicalId":20535,"journal":{"name":"Procedia CIRP","volume":"130 ","pages":"Pages 250-263"},"PeriodicalIF":0.0,"publicationDate":"2024-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142723242","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Image Enhancement for Machine Vision and Industrial Image Processing 用于机器视觉和工业图像处理的图像增强技术

Procedia CIRP

Pub Date : 2024-01-01 DOI: 10.1016/j.procir.2024.10.085

Daniel Weerts , Maren Petersen

Machine vision systems and image processing have become an integral part of today’s production lines. The reasons for this are the high degree of flexibility and adaptability that they offer. However, the robustness of such systems is heavily dependent on stable environmental conditions such as constant lighting. The method presented here is intended to remedy this issue by using a deep learning approach to transfer the characteristics of good images to negatively affected images. In addition to changing light conditions, a possible variety of part colors is also taken into account. The approach is verified using an exemplary pick-and-place application with a smart camera. The experiment resulted in a significant improvement in the object detection task. The smart camera successfully detected objects in images where previous attempts had failed.

机器视觉系统和图像处理已成为当今生产线不可或缺的一部分。原因在于它们具有高度的灵活性和适应性。然而，此类系统的鲁棒性在很大程度上依赖于稳定的环境条件，如持续的照明。本文介绍的方法旨在利用深度学习方法将良好图像的特征转移到受负面影响的图像上，从而解决这一问题。除了不断变化的光照条件外，还考虑了各种可能的零件颜色。使用智能相机的拾放应用程序对该方法进行了验证。实验结果表明，物体检测任务有了明显改善。智能相机成功地检测到了之前尝试失败的图像中的物体。

引用次数: 0

Leveraging Machine Learning for Power Consumption Prediction of Multi-Step Production Processes in Dynamic Electricity Price Environment 利用机器学习预测动态电价环境下多步骤生产流程的耗电量

Procedia CIRP

Pub Date : 2024-01-01 DOI: 10.1016/j.procir.2024.10.080

Muhammad Abdullah Shah , Hendro Wicaksono

Rising energy costs drive a compelling demand for energy-efficient manufacturing across sectors, paralleled by increasing consumer preferences for eco-friendly products. To remain competitive, companies are actively enhancing their energy efficiency. Integrating dynamic pricing in manufacturing, aimed at optimizing renewable energy use, requires strategic adjustments in production planning for sustainability. This research highlights the importance of incorporating dynamic pricing into production planning, emphasizing the need to shift processes to time slots when the energy prices are low or optimal. This study focuses on predicting the power consumption of multi-step CNC machine operations within a production cycle. Utilizing advanced Machine Learning (ML), including neural networks, statistical, and additive models, this research found unique time series characteristics influencing model performance across production steps. A practical use case within a German manufacturing Small and Medium Enterprises (SME) demonstrates how prediction results can optimize production processes in a dynamic pricing environment, providing a blueprint for diverse machinery forecasting models. This research’s insights extend to any industry managing production schedules for multiple machines with various steps in a process cycle. Industries with high energy consumption will benefit significantly through aligning operational efficiency with environmental sustainability goals.

能源成本的不断上涨推动了各行各业对高能效制造业的迫切需求，与此同时，消费者对环保产品的偏好也在不断增加。为了保持竞争力，企业都在积极提高能效。将动态定价纳入制造业，旨在优化可再生能源的使用，这需要对生产规划进行战略性调整，以实现可持续发展。本研究强调了将动态定价纳入生产规划的重要性，强调了将流程转移到能源价格较低或最佳时段的必要性。本研究的重点是预测生产周期内多步数控机床操作的功耗。利用先进的机器学习（ML），包括神经网络、统计和加法模型，本研究发现了影响跨生产步骤模型性能的独特时间序列特征。德国一家中小型制造企业（SME）的实际应用案例展示了预测结果如何在动态定价环境中优化生产流程，为各种机械预测模型提供了蓝图。这项研究的洞察力适用于任何在工艺周期中管理多台机器不同步骤的生产计划的行业。通过将运营效率与环境可持续发展目标相结合，高能耗行业将受益匪浅。

{"title":"Leveraging Machine Learning for Power Consumption Prediction of Multi-Step Production Processes in Dynamic Electricity Price Environment","authors":"Muhammad Abdullah Shah , Hendro Wicaksono","doi":"10.1016/j.procir.2024.10.080","DOIUrl":"10.1016/j.procir.2024.10.080","url":null,"abstract":"<div><div>Rising energy costs drive a compelling demand for energy-efficient manufacturing across sectors, paralleled by increasing consumer preferences for eco-friendly products. To remain competitive, companies are actively enhancing their energy efficiency. Integrating dynamic pricing in manufacturing, aimed at optimizing renewable energy use, requires strategic adjustments in production planning for sustainability. This research highlights the importance of incorporating dynamic pricing into production planning, emphasizing the need to shift processes to time slots when the energy prices are low or optimal. This study focuses on predicting the power consumption of multi-step CNC machine operations within a production cycle. Utilizing advanced Machine Learning (ML), including neural networks, statistical, and additive models, this research found unique time series characteristics influencing model performance across production steps. A practical use case within a German manufacturing Small and Medium Enterprises (SME) demonstrates how prediction results can optimize production processes in a dynamic pricing environment, providing a blueprint for diverse machinery forecasting models. This research’s insights extend to any industry managing production schedules for multiple machines with various steps in a process cycle. Industries with high energy consumption will benefit significantly through aligning operational efficiency with environmental sustainability goals.</div></div>","PeriodicalId":20535,"journal":{"name":"Procedia CIRP","volume":"130 ","pages":"Pages 226-231"},"PeriodicalIF":0.0,"publicationDate":"2024-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142723338","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

The Digital Thread Framework for Implementing Intelligent Machining Applications 实现智能加工应用的数字线程框架

Procedia CIRP

Pub Date : 2024-01-01 DOI: 10.1016/j.procir.2024.10.091

Jeongin Koo , Soohyun Nam , Hoon-Hee Lee , Dong Yoon Lee

The digital transformation of manufacturing industry has led to vast amounts of data, making effective data utilization and analysis crucial for gaining a competitive advantage. However, the diverse formats and complex structures of manufacturing data pose significant challenges to data integration and interoperability. This paper presents a digital thread framework for intelligent machining applications based on a common process planning data model derived from the ISO 14649 standard. The framework integrates various data used in the process planning stage and enables contextual connection of data generated at each stage of the machining process, including virtual machining, machining monitoring, and geometric dimensioning and tolerancing (GD&T). The common data model is constructed by parsing the EXPRESS data model from ISO 14649-1/11 into an OpenAPI Specification JSON format and generating classes in individual programming languages. The digital thread focuses on connecting and restoring the context of operation data, extending the ISO 14649 data model to incorporate tool and equipment information for various applications. The monitoring data is synchronized with the virtual machining data, and the monitoring reference information is mapped to the digital thread project data. The effectiveness of the proposed framework is demonstrated through a reference chattering application, which utilizes parameters from the stability lobe diagram (SLD), machine tool, and virtual machining data. The framework facilitates data analysis and utilization by contextually connecting data generated at each stage of the machining process, ultimately supporting the development of intelligent applications based on monitoring data.

制造业的数字化转型带来了海量数据，有效的数据利用和分析成为获得竞争优势的关键。然而，制造业数据格式多样、结构复杂，给数据集成和互操作性带来了巨大挑战。本文介绍了基于 ISO 14649 标准衍生的通用流程规划数据模型的智能加工应用数字线程框架。该框架集成了工艺规划阶段使用的各种数据，并实现了加工过程各阶段生成的数据的上下文连接，包括虚拟加工、加工监控以及几何尺寸和公差（GD&T）。通用数据模型是通过将 ISO 14649-1/11 中的 EXPRESS 数据模型解析为 OpenAPI Specification JSON 格式并生成各种编程语言的类而构建的。数字线程的重点是连接和恢复运行数据的上下文，扩展 ISO 14649 数据模型，为各种应用纳入工具和设备信息。监控数据与虚拟加工数据同步，监控参考信息映射到数字线程项目数据。通过参考颤振应用，利用稳定叶图（SLD）、机床和虚拟加工数据中的参数，展示了所提框架的有效性。该框架通过将加工过程各阶段生成的数据进行上下文连接，促进了数据分析和利用，最终支持了基于监控数据的智能应用开发。

{"title":"The Digital Thread Framework for Implementing Intelligent Machining Applications","authors":"Jeongin Koo , Soohyun Nam , Hoon-Hee Lee , Dong Yoon Lee","doi":"10.1016/j.procir.2024.10.091","DOIUrl":"10.1016/j.procir.2024.10.091","url":null,"abstract":"<div><div>The digital transformation of manufacturing industry has led to vast amounts of data, making effective data utilization and analysis crucial for gaining a competitive advantage. However, the diverse formats and complex structures of manufacturing data pose significant challenges to data integration and interoperability. This paper presents a digital thread framework for intelligent machining applications based on a common process planning data model derived from the ISO 14649 standard. The framework integrates various data used in the process planning stage and enables contextual connection of data generated at each stage of the machining process, including virtual machining, machining monitoring, and geometric dimensioning and tolerancing (GD&T). The common data model is constructed by parsing the EXPRESS data model from ISO 14649-1/11 into an OpenAPI Specification JSON format and generating classes in individual programming languages. The digital thread focuses on connecting and restoring the context of operation data, extending the ISO 14649 data model to incorporate tool and equipment information for various applications. The monitoring data is synchronized with the virtual machining data, and the monitoring reference information is mapped to the digital thread project data. The effectiveness of the proposed framework is demonstrated through a reference chattering application, which utilizes parameters from the stability lobe diagram (SLD), machine tool, and virtual machining data. The framework facilitates data analysis and utilization by contextually connecting data generated at each stage of the machining process, ultimately supporting the development of intelligent applications based on monitoring data.</div></div>","PeriodicalId":20535,"journal":{"name":"Procedia CIRP","volume":"130 ","pages":"Pages 301-306"},"PeriodicalIF":0.0,"publicationDate":"2024-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142723249","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Efficient Deployment of Machine Learning Models in Manufacturing and Industrial Environments using ROS 使用 ROS 在制造和工业环境中高效部署机器学习模型

Procedia CIRP

Pub Date : 2024-01-01 DOI: 10.1016/j.procir.2024.10.074

Marvin Frisch , Jan Baumgärtner , Imanuel Heider , Alexander Puchta , Jürgen Fleischer

This paper presents a deployment concept that aims to overcome the challenges in the implementation of Machine Learning (ML) models in manufacturing and industrial environments. In these contexts, robots are not typically viewed as production machines. However, the potential for applying advanced techniques such as condition monitoring extends beyond production lines to encompass robotic systems. As a result, there arises a need for a modular solution that integrates into the existing ecosystem while accommodating the requirements of robotic environments. By embracing modularity and interoperability, our proposed deployment concept not only addresses the challenges specific to industrial robotics but also fosters a holistic approach to enhancing operational efficiency and performance in diverse manufacturing settings.

For this, an easily customizable and adjustable system that handles both data acquisition and data transfer is needed. By using the Robot Operating System (ROS) for all necessary data handling, we achieve a highly modular, efficient, and easy-to-use low-code deployment pipeline. Our approach splits the different processing steps into separate nodes and automatically sets up all necessary communication channels, achieving high interchangeability and a quick time-to-deploy. The approach is explained in detail and demonstrated for the real use case of deploying models to monitor handling robots.

本文提出了一种部署概念，旨在克服在制造和工业环境中实施机器学习（ML）模型所面临的挑战。在这些环境中，机器人通常不被视为生产机器。然而，应用先进技术（如状态监测）的潜力已超出生产线，涵盖了机器人系统。因此，需要一种模块化解决方案，既能融入现有生态系统，又能满足机器人环境的要求。通过采用模块化和互操作性，我们提出的部署概念不仅能应对工业机器人技术所特有的挑战，还能在不同的制造环境中促进提高运行效率和性能的整体方法。通过使用机器人操作系统（ROS）进行所有必要的数据处理，我们实现了一个高度模块化、高效且易于使用的低代码部署管道。我们的方法将不同的处理步骤分割成独立的节点，并自动设置所有必要的通信通道，从而实现了高度的互换性和快速部署。我们将详细解释这种方法，并在部署模型以监控搬运机器人的实际使用案例中进行演示。

{"title":"Efficient Deployment of Machine Learning Models in Manufacturing and Industrial Environments using ROS","authors":"Marvin Frisch , Jan Baumgärtner , Imanuel Heider , Alexander Puchta , Jürgen Fleischer","doi":"10.1016/j.procir.2024.10.074","DOIUrl":"10.1016/j.procir.2024.10.074","url":null,"abstract":"<div><div>This paper presents a deployment concept that aims to overcome the challenges in the implementation of Machine Learning (ML) models in manufacturing and industrial environments. In these contexts, robots are not typically viewed as production machines. However, the potential for applying advanced techniques such as condition monitoring extends beyond production lines to encompass robotic systems. As a result, there arises a need for a modular solution that integrates into the existing ecosystem while accommodating the requirements of robotic environments. By embracing modularity and interoperability, our proposed deployment concept not only addresses the challenges specific to industrial robotics but also fosters a holistic approach to enhancing operational efficiency and performance in diverse manufacturing settings.</div><div>For this, an easily customizable and adjustable system that handles both data acquisition and data transfer is needed. By using the Robot Operating System (ROS) for all necessary data handling, we achieve a highly modular, efficient, and easy-to-use low-code deployment pipeline. Our approach splits the different processing steps into separate nodes and automatically sets up all necessary communication channels, achieving high interchangeability and a quick time-to-deploy. The approach is explained in detail and demonstrated for the real use case of deploying models to monitor handling robots.</div></div>","PeriodicalId":20535,"journal":{"name":"Procedia CIRP","volume":"130 ","pages":"Pages 188-193"},"PeriodicalIF":0.0,"publicationDate":"2024-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142723332","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0