共赢行业数智化实践白皮书：拥抱数智化机遇，共赢数智化未来

拥抱数智化机遇，共赢数智化未来 商标声明 ，，是华为技术有限公司的商标或者注册商标，在本手册中以及本手册描述的产品中，出现的其他商标、产品名称、服务名称以及公司名称，由其各自的所有人拥有。 免责声明 本文档可能含有预测信息，包括但不限于有关未来的财务、运营、产品系列、新技术等信息。由于实践中存在很多不确定因素，可能导致实际结果与预测信息有很大的差别。因此，本文档信息仅供参考，不构成任何要约或承诺，华为不对您在本文档基础上做出的任何行为承担责任。华为可能不经通知修改上述信息，恕不另行通知。 目录 序言一 The impressive growth of AI is openingup new prospects for the developmentof intelligent systems, particularlyautonomous systems designed to replacehuman agents in complex organizations,as envisaged by industrial IoT. It shouldbe noted, however, that AI is currentlyfocused on conversational systems thatinteract with a human in the loop. TheAI industrial revolution has only justbegun! Certain technical advances,such as intelligent sensors, prediction/analysis engines and intelligent O&Massistance, have improved the level ofindustrial autonomy. The autonomouscar sector is leading the way with mixedsuccess, while progress in other sectorssuch as avionics, smart factories andtelecommunication networks is moregradual. and the corresponding applications inthousands of industries, the world issteadily moving towards industry digitalintelligence. asevidenced by the organizat ionofinternat ional summits and themobilization of international institutionssuch as the UN. to actively contribute to research andinnovation in cutting-edge technologies,to bridge the gap and achieve digitalintelligence in all industrial sectors and,ultimately, to successfully meet globalchallenges and contribute to the well-being of humanity. It is already clear that the industrialrevolution expected from AI will nothappen without a substantial effort tobridge the gap between well-establishedi n d u s t r i a ls y s t e m s e n g i n e e r i n gpractices, which follow standards-baseddevelopment workflows and aim fortrustworthiness guarantees, and AIsystem. This requires a collective effortto develop a common domain-specificdata infrastructure and technology,as well as sorely lacking methods forbuilding accurate AI models that arestatistically consistent with their trainingdatasets. Better coordination of thiseffort needs to be organized aroundreference architectures and standardsfor each domain, which enable differentsolutions to be compared and assessed.This requires international consultationand coordination between countrieswith a strong and extensive industrialbase, with leading industries playing akey role. I am convinced that all countries mustmobilize collectively and in solidarity Joseph Sifakis, Turing Award 2007 The advent of the digital and intelligentera raises enormous technical andscientific challenges, not least that ofmarrying AI with traditional digitalsystems engineering, as AI systemsneed to be integrated into industrialsystems for which we have determinedreliability guarantee requirements. Awell-known limitation of AI systemsis that it is difficult to guarantee theirsafety and limit the risks associated withtheir use. AI safety is a global concern, This white paper shows that withthe innovation and development ofcutting-edge technologies such as AI chains. Second, digital and intelligenttransformation is of strategic importanceand, thus, requires a strategic plan,e.g. in form of roadmaps and maturitymodels. Both would provide guidanceand means of coordination to businessesand international activities, e.g. toreach the sustainable developmentgoals. Third, as data is the strategicresource of the digital era, the datavalue chain, i.e. the entirety of activitiesto create, collect, pre-process, store,make accessible and share data, shouldfunction as the conceptual foundationto deploy various digital technologies. Digital and intelligent technologiesare key to seize business opportunitiesin almost all industrial sectors andto contribute to the accomplishmentof globally shared goals such as theSustainable Development Goals targetedby the United Nations. been achieved when it comes to digital and intelligent transformation and thatresearch and technology development provides us with the means to even accelerateour transformation path – for the better of societies and businesses all over theworld. Prof. Dr.-Ing. Boris Otto, Institute Director at the Fraunhofer Institute for Software andSystems Engineering ISST The advancement of technologicalinnovation in the digital and intelligentspace is immense and is continuingto accelerate. Data sharing platforms,the transformation to an edge-cloudcontinuum and the power of artificialintelligence are prominent examples. However, to leverage the potentialof digital and intelligent innovationt ot h e m a x i m u m e x t e n t p o s s i b l e ,three points are pivotal for success.First, technologies such as IoT, bigdata, artificial intelligence, and cloudcomputing do not form isolated sourcesof progress but must be developedaccording to an integrated architecturestack. A reference intelligent architecturewould function as a blueprint forseveral activities which require globalcollaboration such as standardizationand the suppor