Deep Dive: Technology Directions - Building the Digital Factory Piece by Piece

AI权益加码!Claude Code、Cursor等20+工具免费用! 购周边限时加赠Coding Plan Lite,畅享主流AI工具!学习进阶更高效! 阅读详情

http://www.managingautomation.com/maonline/magazine/read/view/Deep_Dive_The_Digital_Factory_Is_a_Work_in_Progress_244432?page=6

 

Given the worldwide economic recession, manufacturers are under pressure to become more resourceful and agile. As a result, companies are starting to take a serious look at the value of incorporating more simulation and visualization into design-to-production processes so that they can see and test production setups before actually making new production investments.

Technology vendors are taking note and introducing products that move manufacturing closer to the virtual reality of the digital factory.

While industry experts say we are still some years away from technology tools that support a soup-to-nuts digital manufacturing environment, many vendors are coming up with point solutions that offer ways to close the loop between the design of a product and actual production. Vendors are finding, however, that integrating those point solutions into a unified whole is a challenge, in part because the software programs that drive the different domains — CAD and control — usually haven't been designed to communicate with one another. In addition, an integrated digital factory environment must be able to manage the information that's generated by the different applications involved. That's why product lifecycle management (PLM) is an important application to include in the digital factory discussion.

Automation and PLM leader Siemens is one vendor that is going back to the drawing board to create a turnkey solution that incorporates all of these technologies in a tightly integrated offering. Meanwhile PLM software provider Dassault Systèmes, which has partnerships with Rockwell Automation, Omron Corp., and Schneider Electric on the control side, as well as iBASEt and Intercim on the manufacturing operations management side, is offering manufacturers multiple digital factory entry points.

In addition, Dassault and other vendors, such as Autodesk and Invensys Process Systems, offer visualization tools that virtualize the factory floor experience.

The digital factory concept, at its core, is about modeling and integrating design and production processes. The challenge, however, is keeping digital models in sync with the physical world, particularly the plant floor, where change is constant. Often, as soon as the physical plant is running, the digital model diverges from what's going on in the real world, says Patrick Michel, Dassault's vice president of Delmia industry solutions and marketing. "Everyone is working to extend [the digital model] to have a permanent view of the accurate state of what is being built, as well as to incorporate knowledge sharing [between engineering and manufacturing]," he says.

The work being done by Delmia and Intercim, for example, will tightly integrate Dassault's V6 PLM architecture, specifically its Delmia digital manufacturing software — which includes early process planning, modeling, and simulation of assembly lines — with Intercim's Pertinence software suite for executing processes, tracking workflow, and managing quality during production.

The goal is to align manufacturing with engineering and PLM, the companies say. Specifically, that means creating an "as built" model that can tie back to the "as designed" model using a common language.

"You can see a virtual representation of the factory, take a snapshot of that, and simulate scenario changes before spending any money," says Judson Plapp, Intercim's vice president of marketing and corporate strategy. "That's the point of closing the loop. Getting real data and feeding that back to engineering."

Dassault made a minority investment in Intercim in order to facilitate the sharing of intellectual property (IP) and to ensure tight integration between the two companies' technologies. But Dassault is also working with iBASEt on an integrated offering — absent any kind of financial investment.

Both iBASEt and Intercim cater to the aerospace and defense industry. In aerospace, work instructions must include process definitions. There is a desire to automate work instruction definition as much as possible, says Conrad Leiva, vice president of product marketing and alliances at iBASEt. Simulation and animation are becoming integral to managing production processes in this industry.

But even this must be approached cautiously. "The expectation is to leverage simulation, but, at the same time, you don't want to simulate everything. Some things are simple and intuitive, and you don't need to simulate it," as simulation adds more cost into the equation, Leiva says.

Sensible Simulation

Simulation does not have to be applied everywhere, but there are certain areas of the design-to-manufacturing process where it can simplify the overall effort.

For example, in another attempt to tie together equipment control and engineering, Rockwell Automation teamed up with Dassault's Delmia over a year ago, and today Rockwell sells a utility, called RSLogix 5000 version 17, that creates bidirectional synchronization between Rockwell control systems and the Delmia digital manufacturing software. Together, the technologies act as a framework for mapping out production systems, saving time in production planning. But the utility also improves quality by simulating what-if scenarios before a product rollout.

"Up until this point, systems were integrated using tools to build 3D models, and it was dumped into the control system," says Jason Weber, strategic alliance manager at Rockwell Automation. "It saves time initially, but as soon as it's done, it's broken because there is no link back and forth. If something changes on the mechanical side, you have to send an e-mail or walk over to make sure everyone knows what the changes are," he says.

This wastes time, Weber says. But by synchronizing mechanical engineering and control information in a common object database, there is a bidirectional communication between systems that updates all files immediately when changes occur. "For end users, there is less error as you get the machine or work cell up and running. That is a big win in efficiency and time to production," he says.

For its part, Siemens two years ago acquired UGS, resulting in what is now called the Siemens PLM business unit. Currently, work is ongoing behind the scenes at Siemens on a project, dubbed Archimedes, that will integrate all of the aspects that Dassault is attempting with multiple partners, but it will be under one Siemens-based architecture, using the company's Teamcenter PLM as the data management platform.

Project Archimedes — which is still under wraps at Siemens — could take a while to become a reality, industry observers say. But manufacturers don't have to wait to reap the benefits, as there are aspects of the digital factory that exist in the Siemens product portfolio today.

For example, Tecnomatix is branded as the company's "digital manufacturing solution." The Tecnomatix strategy started in 1983, when the company developed software initially intended to help BMW speed up the planning of its production lines. Since then, Tecnomatix has added simulation for verifying processes and is able to transfer the processes to the shop floor. Tecnomatix, however, does not complete the feedback loop that ties the production lifecycle back to design and PLM. This adaptive approach — from concept, to design, to automation, and back —requires integration among Siemens CAD/CAE, simulation, PLM, MES, and control architectures.

Nevertheless, the benefits of Tecnomatix alone are significant.

"It makes planning more efficient, ramp-up shorter, safer, and we are also providing a lot of technology to optimize the manufacturing processes on the lines by increasing the throughput of the lines, [but] in a way that minimizes the investment," says Stefan Linner, vice president of Tecnomatix product marketing at Siemens PLM Software.

To that end, there are many examples of customers saving millions of dollars as a result of using the Tecnomatix software, he says. "Digital manufacturing is a reality, and many customers out there are getting a lot of value [out of what's available today]," he says.

Factory Animation

A good starting point for manufacturers wanting to get their digital feet wet would be to map out the factory floor. Autodesk's NavisWorks software, for example, takes a digital prototype of a product and extends it into a virtual prototype of production, accounting for collision detection and clearance analysis as a product moves down the line.

"It is not at the level of doing PLC and robotic validation yet," says Hilde Sevens, senior product manager of the Autodesk Manufacturing CIM group. "But it does factory layout and optimization." And it is easy enough for the casual user to use, she says.

Similarly, Next View Software Inc. offers a data-driven visualization tool for managing warehouses or yards. including material handling and labor management. Based on Microsoft Corp.'s XNA framework, which is the foundation of Xbox, Next View created a tool that can be put into the hands of a kid just coming out of high school.

"We believe software applications of the future will have to look, feel, and behave different than they have in the past," says Steve Simmerman, head of business development at Next View.

The company's software takes real-time feeds from the warehouse to show what is happening in the facility. "We build a scale model based on XYZ coordinates ... that can take a bird's eye view as you fly into the facility or dissolve ceilings and walls and let you see the forklift people moving in real time," he says.

The system takes information from labor management systems in order to identify who is logged in and what equipment they are using. It can also track the material flow based on purchase orders from an ERP system. In addition, it can tie into transportation networks and MES software to show the status of manufacturing.

The Next View visualization software can be used to simulate the build-out of a facility. "If you want to double the size of the warehouse, our modeling software can take real data and assess the impact on the facility," Simmerman says.

While this kind of real-time visualization has been used in discrete industries, where packaged goods come off a line, Invensys Process Systems in February unveiled its own virtual reality technology for the oil and gas and chemical industries.

Invensys' Immersive Virtual Reality Process Technology is a next-generation human-machine interface (HMI) tool that couples dynamic simulation software with a high-fidelity, 3D model of the plant that is enabled by a virtual reality engine. The tool is primarily used for training operators. Users wear a stereoscopic headset, immersing them in an environment that allows them to move freely through the plant and see all of the equipment and instrumentation as it appears in real life. Complex processes can be simulated to teach an operator how to take down a plant, for example, in case of an emergency.

The technology, which will be commercially available in the second half this year, also has the potential to link to actual plant operations. "In the future of HMI, for example, the virtual reality environment may not be linking only to a simulation system, but also to real-life data," says Maurizio Rovaglio, director, Invensys global consulting. But he concedes that is still a futuristic concept.

Indeed, today there is much about the digital factory concept that is futuristic. But, while vendors have yet to provide a set of fully integrated tools that can support every aspect of the digital factory, their point products take much of the complexity out of directly integrating product and production models.

MCP + 应用生成:让 AI 直接产出可交互的应用 2025年,MCP迎来革命性升级:可交互输出让AI不再仅返回文本,而是生成带表单、图表、按钮的结构化UI卡片。用户可在卡片中直接筛选、导出,实现“一问一操作”闭环。通过返回UI Schema或独立HTML片段,结合客户端渲染与后端调用回流,显著提升效率(任务轮次↓65%),尤其适合数据看板、表单填报等重操作场景。但需克制:纯查询仍以文本为佳。落地关键在于合理设计交互、控制复杂度,并规避版本不一致、数据过载等六大坑。本质是将AI从“告知者”变为“应用交付者”,开启智能体原生交互新范式。 阅读详情

相关推荐

SSP-354-Jetta2006-捷达2006版自学手册

内容概要:本文档为大众捷达2006款的自学手册,全面介绍了该车型的设计、技术参数及各项系统配置。捷达2006在安全性、舒适性、操控性和空间布局方面均表现出色,采用高强度车身结构与激光焊接技术提升刚性,并配备先进的主被动安全系统,如多气囊、ESP MK60、ISOFIX接口等。车辆提供多种发动机选择,涵盖1.6L至2.0L汽油及柴油动力,部分搭载涡轮增压与直喷技术,满足EU4排放标准。传动系统包括手动、自动及DSG双离合变速箱。内饰设计注重实用性,拥有丰富的储物空间和可选装的Climatronic双区空调、电动座椅等功能。电子架构采用多总线网络系统,集成动力、舒适、信息娱乐等多个控制单元,支持诊断通信与智能功能扩展。; 适合人群:汽车维修技术人员、汽车工程专业学生以及对大众捷达2006款车型结构和技术细节感兴趣的车主或爱好者。; 使用场景及目标:①用于深入理解捷达2006的整车构造与各子系统工作原理;②作为维修、保养和故障诊断的技术参考资料;③辅助进行汽车电子系统学习与实操培训; 阅读建议:本手册内容专业性强,建议结合实物车辆或相关教学设备对照学习,重点关注各系统的结构图示、技术参数表及控制网络拓扑,以便更好地掌握整车技术细节。

微软技术日报 2026-09-08:Patch Tuesday 任务栏自由了,GPT-6 Astra 四平台齐发

今天推送的 9 月补丁让 Windows 11 找回了任务栏与开始菜单的自由,而 Copilot 一周内接入两大前沿模型、Studio Harness 转 GA,则标志着微软的"智能体平台化"从口号进入账单期——本地跑大模型(Project Zenith)与云端多模型编排(Copilot Studio)两条路线正在同时加速。

海盗Sharp的博客 345

【物联网开发】基于DevEco与MQTT的北向应用快速开发:华为云IoTDA平台设备监护系统设计

内容概要:本文介绍了基于华为云IoTDA平台的北向模拟环境监护快速应用开发全流程,涵盖从开发环境搭建、北向应用开发、产品与设备模型构建,到数据采集控制、命令下发及南北向系统联调的完整实践。通过DevEco Studio进行前端界面开发,利用MQTT.fx模拟设备通信,结合ArkTS语言实现认证鉴权、设备列表展示、设备详情查看等功能,并通过HTTP接口与IoTDA平台交互实现设备管理、数据查询与命令下发。同时,文档还涉及南向硬件HiSpark开发板的应用编译与烧录流程,形成端到边到云的闭环开发示范。 适合人群:具备一定前端开发基础和物联网基础知识,熟悉TypeScript或类Web开发框架,从事物联网应用开发1-3年的研发人员。 使用场景及目标:①快速构建面向华为云IoTDA平台的北向监控类应用原型;②掌握设备接入、数据上报、指令下发等核心业务流程的实现方法;③实现前后端分离架构下的设备可视化管理与远程控制功能。 阅读建议:建议结合华为云IoTDA控制台实际操作,同步搭建DevEco Studio与MQTT.fx环境,逐步实现文档中的各模块功能,重点关注HTTP请求封装、路由跳转、状态管理与JSON数据解析等关键代码实现。

微软技术日报 2026-09-09:Project Opal 自主智能体亮相,Patch Tuesday 966 漏洞含 2 零日

每天 5 分钟,速览微软技术圈最新动态。今日看点:Project Opal 让 Copilot 学会"自主干活"、创纪录的 966 漏洞 Patch Tuesday、WinUI 3 三十分钟速成新范式。

海盗Sharp的博客 266

GitHub今日热榜 | 2026-09-08:微软 markitdown 冲进前三

<think>好的,用户需要根据提供的GitHub趋势榜内容,写一篇不超过150字的文章摘要。内容包含榜单变化、亮点项目和分析。需要简洁覆盖核心要点:今日榜单变化大,新进8个,两个持续项目撑住,微软markitdown和字节deer-flow等新面孔上榜,主题是AI Agent工具链。控制字数。 先梳理关键信息:今日榜单10个中8个是新进,昨日霸榜项目集体跌出,ECC和FckSignups持续。微软markitdown冲至第二,字节deer-flow等新项目上榜,主线是AI Agent技能与工具链。需要把这

m0_68631449的博客 373

微软技术日报 2026-09-06:微软天顶计划定制开发者 Win11、Aspire 全面接入 AI 智能体

覆盖 Windows、开发者生态、.NET/开发工具、Azure 云、安全公告五大板块,聚焦技术动态,不含财务信息。

海盗Sharp的博客 484

python进阶(高阶函数,特性梳理)

高阶函数 高阶函数英文叫Higher-order function。什么是高阶函数?我们以实际代码为例子,一步一步深入概念。

2601_96229461的博客 160

微软技术日报 2026-09-07:Win11 默认开启内存完整性、统一版 Copilot 桌面端在即、明日 Patch Tuesday

覆盖 Windows、Copilot/AI、Azure 云、开发工具、安全公告与硬件六大板块,聚焦技术动态,不含财务信息。

海盗Sharp的博客 308

AI Agents for Beginners:微软官方 AI Agent 入门课程,18 课掌握智能体开发

Microsoft Cloud Advocates 团队开源的一套免费、系统性的 AI Agent 入门课程。——18 个课时,从 AI Agent 基础到完整的应用开发,手把手带你入门。它瞄准的是现有学习资源的痛点:AI Agent 是生成式 AI 的最新前沿,但大多数教程要么过于理论、要么只讲零散技巧,缺乏系统性;而官方文档又往往不够友好。

风雪夜归人的博客 241

Generative AI for Beginners:微软官方 21 课生成式 AI 入门教程,从零掌握 AI 应用开发

Microsoft Cloud Advocates 团队开源的一套免费、系统性的生成式 AI 入门课程。——21 个课时,从生成式 AI 基础到完整的应用开发,手把手带你入门。它瞄准的是现有学习资源的痛点:生成式 AI 领域知识更新极快,很多教程要么过于理论、要么只讲零散技巧,缺乏系统性;而官方文档又往往不够友好。这套课程由微软 Cloud Advocates 精心设计,每节课都有视频讲解、代码示例(Python + TypeScript 双语言)、额外学习资源,是真正从零开始的系统性学习路径。

风雪夜归人的博客 239

Windows11自定义分区大小安装系统

格式化为NTFS,标签为Recovery。设置恢复分区磁盘ID,这一步在前面给出的微软官方有说明,EFI分区和MSR分区不需要设置ID,是因为create命令中已经给出了对应的分区类型,但由于恢复分区和主分区同时使用NTFS,也没有单独的名称(Recovery是给用户看的,不是给机器识别的。就像电脑中的C盘可能对外名称是本地磁盘,但是机器读的是路径C:)如果没有给恢复分区设置ID,则会创建一个基本数据分区,便于理解的说法,是会创建一个D盘,而非恢复分区。

问道至简 178

CSharp: Task Scheduler

本项目为基于C# .NET 10的定时任务调度系统,采用模块化设计,支持YAML配置、SQLite存储、多数据库兼容。核心功能包括:基于cron表达式的精准调度(自研解析器)、并发控制、任务重试、日志滚动(按5MB切割),支持命令行与GUI双界面操作。提供任务增删改查、一键执行、导入导出JSON、实时仪表盘统计等功能,具备一次性任务自动禁用、执行历史追踪等特性,适用于Windows环境下的自动化运维场景。

Geovin Du Dream Park 157

微软技术日报 2026-09-10:Project Zenith 开发者设备亮相,.NET 11 RC1 可投产

今天的微软把重心放在「开发者的一公里」:设备端(Project Zenith)、运行时(.NET 11 RC1)、模型层(Copilot 多模型矩阵)三层同时发力,开发者环境从"自己搭"走向"买来就能写"。参考来源:微软官方博客(Project Zenith)、.NET 官方发布说明、TechRepublic、腾讯新闻科技频道、第一财经、IT之家。本文由 AI 辅助整理,仅覆盖技术动态,不含财务信息。

海盗Sharp的博客 364

计算机原理—程序的具体执行流程

<think>我们根据内容生成摘要,≤150字。需要覆盖:程序加载后并非直接main,Linux/Windows都经动态链接器/加载器、CRT初始化后调用main,退出时资源回收。字数限制。</think>程序加载后并非直接进入main,而是由操作系统创建进程,经动态链接器/加载器完成资源、依赖、重定位等初始化,再由C/C++运行库(如_start或mainCRTStartup)构造全局对象后调用main;退出时经标准退出流程回收资源。Linux与Windows机制相似,但具体实现不同。

fpcc的专栏 543

【LibreCAD 2D架构】从两个坐标到图形实体:RS_ActionDrawLine如何创建RS_Line

本文以 LibreCAD 的 RS_ActionDrawLine 为切入点,解析从鼠标点击到图形创建的完整交互流程。通过梳理鼠标事件、坐标转换、捕捉机制、状态机、预览绘制及正式实体创建等关键环节,说明用户操作如何逐步转化为 CAD 图形。同时结合输入层、交互层、坐标层、几何层、数据管理层和显示层,揭示 LibreCAD 清晰的分层设计思想,帮助读者理解绘图功能背后的源码组织与运行机制。

🐣 一名正在努力升级的 C/C++ 程序员、喜欢研究源码、拆解架构,也喜欢把复杂的问题讲清楚。正在学习 Linux、C++、图形软件开发等。不一定每次都能写出完美的代码,但会认真搞懂每一个问题。 2448

GPT-6 + Blender 完整教程:用官方 API 给 Blender 接一个 AI 脚本助手

GPT-6 Astra 是 OpenAI 在 2026 年公开的旗舰模型,Blender 是可用 Python 扩展的 3D 创作环境。把两者接起来,最稳的方式不是直接操控界面,而是让 GPT-6 先生成可审查的 Blender 脚本,再由插件写入文本块或执行。官方模型页给出 1,050,000 token 上下文和 128,000 token 最大输出,足够支撑复杂建模脚本与分步修正。

aidoudoulong的博客 721

微软技术周报 2026-08-31~09-07:GPT-6 Astra 登陆 Foundry、M365 全球认证故障复盘、C# 15 封闭层级

本期聚焦技术动态,不含财务信息。信息来源:Microsoft Learn、.NET News Daily、Azure Weekly(Issue 578)、Microsoft Security Blog、MSRC、Pureinfotech、9to5Windows、Neowin、BleepingComputer、InfoQ 中文、cnBeta、快科技、Microsoft 365 Message Center 等中英文媒体与官方渠道。

海盗Sharp的博客 427

Agent开发 - 实现人类与Manus智能体的终端窗口命令交互

<think>我们根据要求生成≤150字的文章摘要。内容是关于openManus添加askUser工具实现人机交互。摘要要点:核心思路、askUser工具、prompt引导、配置开关、ApplicationRunner实现启动交互。字数限制150字,需简洁。</think>本文介绍了OpenManus人机交互的实现方法:通过添加askUser工具暂停Agent等待用户输入,在SystemPrompt中设定必须询问用户的场景,利用配置开关控制交互模式,并通过实现ApplicationR

weixin_45839019的博客 296

微软技术日报 2026-09-11:VS Code 语音与定时自动化上线,Copilot 键终于可改映射

每天 8 点,5 分钟看懂微软技术圈。今天是 2026 年 9 月 11 日,星期五。

海盗Sharp的博客 238
上一篇: 解析ABB收购著名软件厂商Ventyx的意义与影响
下一篇: Siemens PLM Unveils Context-Driven Collaboration Strategy
lybra
博客等级 码龄20年 41粉丝 100原创
评论
成就一亿技术人!
拼手气红包6.0元
还能输入1000个字符
 
 条评论被折叠 查看
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值