制造业是美国经济的重要组成部分，占国内生产总值（GDP）的12.5%，并提供了1740万个岗位。维持一个强大的工业部分是确保我们未来竞争力之关键，更好地利用数据和分析是建立一个健全的制造业之关键。重要的全国性项目也认同此观点，例如奥巴马政府建立的“国家制造业创新网络（NNMI）”这个开发先进制造科技的研究机构的骨干群体。根据来自市场情报公司IDC的2014年估算，这种被称之为“智能制造”的方式将在未来四年中创造3710亿美元全球净价值。它也能够通过设计过程的精益化（流线化），工厂运营的提升，以及供应链中的风险管理这三种主要途径使美国制造商在全球经济中具有竞争力。
数据能够从产品生命周期的开端就支持制造创新。数据驱动的设计能够在制造任何实物产品之前预告设计过程，从而降低成本，并确保最终产品更加符合客户偏好。例如，欧特克（Autodesk）的计算机辅助设计软件虽已流行了几十年，而现在该公司在其建立的算法生成设计（algorithmically generated design）的研究项目“追梦计划”中，正在开发一种更加数据驱动的方式。这项成就最终将与公司的其他产品整合，使设计者能够根据一系列材料和性能要求生成设计，随后这些设计将被用来进行高精度增材制造。汽车制造商，如沃尔沃（Volvo）和迈凯轮（McLaren）都采取了基于模拟的方式。迈凯轮在制造实体原型之前，对其设计进行性能分析；沃尔沃则整合客户数据来预测一个特定设计或功能是否能够吸引客户。数据驱动的设计在其他领域（如网页设计）已经成为标准做法。制造商若将其更全面地应用于自己的产品，将会受益匪浅。
制造商还可以利用数据和分析来提高工厂车间操作。低成本传感器技术（即物联网）的扩张已经使得几乎所有制造工艺和零部件成为潜在数据源。创新的制造商可以利用所得到的数据集来深入了解实体制造过程，以提高效率，增加产量，并降低产品缺陷。雷神（Raytheon）公司非常在行追踪工厂中一颗螺丝转动的次数，其他公司也正在尽可能多地收集自己的流程细节。例如哈雷戴维森（Harley Davidson）追踪在其摩托车喷漆区域的风扇转速，并能够根据环境波动来通过算法调整风扇。默克（Merck）提高其一种疫苗质量的方式是通过150亿次的计算来决定影响最终产品质量的环境和流程因素。英特尔（Intel）利用数据的预测模型来预测故障，优先检查，并削减其芯片制造工厂的监控成本。 有如此多潜在变量可追踪，“信息太少”不应再是工厂环境中浪费和损失的借口。
最后，数据分析可以帮助制造商管理供应链。产业供应链的相互关联性使其成为风险的温床，并且更多信息意味着返工和成功出货的区别。 GE石油和天然气现在使用的是一个基于云的供应链数据平台来管理他的材料、设备和服务 。这个实时系统现在部署在五个大洲，是为了抗衡油田停工期的高成本。惠普（HP）将网络分析整合入其供应链的监控。他们的做法，其中还包括数据可视化，已经将供应链优化项目所需的时间降低了50%。（美国）国家标准与技术研究院（NIST）正在进行一个项目为“制造数据分析”制定标准、方法和协议，其中一个主要动机是日益增长地对更加全面的供应链智能化的需求。即使有数据驱动的工厂和设计工作室，制造商也只有当其供应链与监控相同水平时才能高枕无忧。
数据与分析能够在流程中的几乎每一个环节帮助制造商，从他们的全球供应足迹到工厂中一颗螺丝的转动。而正由于一个健全的制造业是一个健全的经济体系的重要组成部分，数据驱动的制造业的优势也将遍及全国。
英文原文：
Data is the Key to the Factory of the Future
Manufacturing is a major part ofthe U.S. economy, responsible for 12.5 percent of gross domestic product (GDP)and supporting 17.4 million U.S. jobs. Maintaining a strongindustrial sector is critical to ensuring our future competitiveness, andmaking better use of data and analytics is key to building a healthy manufacturingindustry. Major nationwide initiatives such as the Obama Administration’s NationalNetwork for Manufacturing Innovation (NNMI), a cadre of researchinstitutes to develop advanced manufacturing technologies, have acknowledgedjust that. “Smart manufacturing,” as this approach is often called,could create $371 billion in net global value over the next four years,according to a 2014 estimate from the market intelligence firm IDC. Itcould also help make U.S. manufacturers competitive in the global economy inthree main ways: streamlining the design process, improving factory operations,and managing risk in the supply chain.
Data can support manufacturinginnovation from the very beginning of a product’s lifecycle. By helping inform thedesign process before any physical product is created, data-driven design cancut costs and ensure that final products will be better aligned with customerpreferences. For example, Autodesk’s computer-aided design software has beenpopular for decades, but the company is now working on an even more data-drivenapproach with its algorithmically generated design research initiative ProjectDreamcatcher. That effort, which will eventually be integrated into thecompany’s other products, allows designers to generate designs based on a listof material and performance requirements, which can then be additivelymanufactured with a high degree of precision. Vehicle manufacturers such asVolvo and McLaren have taken a simulation-based approach, withMcLaren conducting performance analysis on its designs before creating a physicalprototype, and Volvo integrating customer data to forecast whether a particulardesign or feature will appeal to its customers. Data-driven design has become a standard practice in other fields such as web design.Manufacturers stand to benefit if they apply it more comprehensively to theirown products.
Manufacturers can also use dataand analytics to improve operations on the factory floor. The explosion of low-cost sensor technologies (i.e. the Internet of Things) has made nearly everymanufacturing process and component a potential data source. Innovativemanufacturers can use the resulting data sets to gain insights about thephysical fabrication process, improving efficiency, increasing yields, andreducing product defects. Raytheon famously keeps track of how many times a screw has been turned in its factories, and othercompanies are working to collect as much detail about their own processes.Harley Davidson, for example tracks fan speed in its motorcycle painting areas and can algorithmicallyadjust the fans based on environmental fluctuations. Merck improved one of its vaccines by conducting 15 billioncalculations to determine what environmental and process factors influenced thequality of the final product. Intel uses predictive modeling on data to anticipatefailures, prioritize inspections, and cut monitoring costs at itschip manufacturing plants. With so many potential variables to track, no longershould “too little information” be an excuse for waste and loss in the factoryenvironment.
Finally, data analytics can helpmanufacturers manage their supply chains. The interconnected nature ofindustrial supply chains makes them hotbeds for risk, and more information canmean the difference between a recall and a successful shipment. GE Oil and Gasnow uses a cloud-based supply chain data platform to manage its materials,equipment, and services. The real-time system, now deployed on five continents,was created to contend with the high cost of downtime at oil fields. HP integrates network analysis into its supply chain monitoring. Their approach, which alsoincludes data visualization, has reduced the time required for supply chainoptimization projects by up to 50 percent. The National Institute of Standardsand Technology (NIST) is working on a project to develop standards, methods, and protocols for manufacturing dataanalytics, a key motivation for which is the increasing demand formore comprehensive supply chain intelligence. Even with data-driven factoriesand design studios, manufacturers will not be able to rest easy until theirsupply chains are subject to the same level of monitoring.
Data and analyticscan help manufacturers at nearly every step in the process, from their globalsupply footprint to the turn of a screw in their factories. And because ahealthier manufacturing sector is an important part of a healthier economy, thebenefits of data-driven manufacturing will be felt throughout the country aswell.
源地址：http://www.datainnovation.org/2014/10/data-is-the-key-to-the-factory-of-the-future/?utm_source=feedly&utm;_reader=feedly&utm;_medium=rss&utm;_campaign=data-is-the-key-to-the-factory-of-the-future
原文发布时间为：2014-10-18