1、Accelerate,innovate,empowerREPORTHow industry leaders leverage AI to boost next-generation manufacturing and operationsPrefaceIn todays rapidly transforming business landscape,the integration of artificial intelligence(AI)in operations has become not just a strategic advantage,but a necessity for re

2、maining competitive.AI has become a transformative force,offering unparalleled opportunities to optimize processes,enhance decision-making and drive efficiencies across the value chain.As industries grapple with challenges such as shortening innovation cycles,sustainability requirements and labor sc

3、arcity,the need to adopt AI solutions and the opportunity to benefit from technology have never been more apparent.But to maximize its potential,the technology needs to be applied in practical use cases where clear value is created.This whitepaper explores the current landscape of tangible AI use ca

4、ses in operations,drawing from a diverse field of applications showcased in past and present Microsoft Intelligent Manufacturing Awards(MIMAs).The awards,now in their fifth year,celebrate the most pioneering digital operations solutions from across industrial sectors,rewarding companies Cover:Westen

5、d61/Getty Images2 Roland Berger,Microsoft,NVIDIA|Accelerate,innovate,empowerthat apply real-life solutions to leverage efficiencies and sustainability.Hosted by Roland Berger and Microsoft,entrants are judged by a high-profile jury from industry and academia,this year under the motto accelerate,inno

6、vate,empower.This paper presents seven key factors common to companies that have successfully adopted AI in operations,using MIMA use cases as examples.These aim to provide practical guidance to business leaders and strategists on navigating the process of AI adoption namely,a clear strategy,transpa

7、rency on existing and required capabilities,and a laser-like focus on the targeted added value.Rupert StuetzleGeneral Manager Manufacturing&Mobility EMEA,Microsoft Marcus Berret Global Managing Director,Roland Berger3 Roland Berger,Microsoft,NVIDIA|Accelerate,innovate,empowerManagement summaryThe us

8、e of AI in operations promises significant benefits.But adoption remains challenging for businesses due to barriers such as proving business value,missing infrastructure and a shortage of skilled AI talent.To overcome such obstacles and embrace intelligent next-generation manufacturing and operation

9、s,companies should follow seven key success factors common to industrial digital leaders:1 Develop a clear AI roadmap,including explicit objectives,executive sponsorship and a clear organizational change process2 Determine the technical foundation,keeping data secure while enabling accessibility and

10、 rapid solution scaling3 Embrace modular platforms to simplify scaling,adaptation to future products and maintenance4 Leverage off-the-shelf technology solutions and enable a fail fast approach5 Democratize digitalization,enabling specialists across hierarchy levels to turn domain knowledge into int

11、elligent solutions6 Empower workforces with AI-based solutions to overcome labor shortages7 Exploit AIs potential to disrupt value creation by using AI-generated insights to improve core offering,maintain installed base and create new products.Roland Berger and Microsoft can support companies throug

12、hout their AI-in-operations journeys.RB N3XT is a dedicated team of specialists for strategic digitalization:rich expertise,proprietary toolkits and proven strategic approaches allow RB to support clients in solution identification,smooth adoption,and maximization of potential.Microsoft supports ind

13、ustrial clients with a broad set of modular offerings.For example,customers can harness Copilot Microsofts generative AI work companion in Microsoft products they already use or learn how to build and modernize AI-powered applications by leveraging Azure OpenAI Service.Clients can build custom AI as

14、sistants to transform their operations and drive innovation across the value chain.To accelerate time-to-value,Azure Marketplace contains off-the-shelf partner solutions for direct deployment.4 Roland Berger,Microsoft,NVIDIA|Accelerate,innovate,empowerContents 1 2 3 45AI in operations:The benefits a

15、nd challenges Key success factors for AI in operations Outlook:The future implications of AINVIDIA:An expert perspective on the future of AI in operationsHow we can support AI in Operations610171819Page5 Roland Berger,Microsoft,NVIDIA|Accelerate,innovate,empower1AI in operations:The benefits and cha

16、llenges AI is key to maintaining competitiveness,enhancing sustainability and tackling labor shortages in operations but adoption requires planning and overcoming obstacles Today,artificial intelligence is everywhere,be it managing smartphone messages,running a businesss chatbot or controlling an in

17、dustrial machine.Use cases for the fast-evolving technology are near limitless,with each highlighting AIs tremendous potential to disrupt all aspects of our lives.The industrial and operations space is no exception.Indeed,use cases have been demonstrated all along the value chain.A THE BENEFITSThe b

18、enefits of AI in operations promise to be significant.For example,R&D experts estimate that conventional AI and generative AI1 applications can deliver up to 60%reduction in product development times,boosting R&D output dramatically in the future.In addition,procurement is set to receive a GenAI-pow

19、ered performance boost of up to 50%improved FTE efficiency,as well as four-times-faster procurement processes.These gains will stem from,for example,improved supplier market screening times,radically parallelized supplier negotiations or significantly improved processing times for contract authoring

20、.B THE CHALLENGESHowever,while AI technologies and research are rapidly advancing,industry is struggling to keep up businesses cannot be expected to turn the technological possibilities into impactful,scaled-up use cases at the same pace.World Economic Forum figures reveal that only one in six manuf

21、acturers have so far successfully incorporated AI into their operations.1 GenAI:deep learning models that can generate high-quality text,images and other content based on the data they were trained on The industry is in arrears with its adoption rate for AI,considering the greater-than-ever pressure

22、 on productivity and efficiency goals,compliance with sustainability targets,and increasingly scarce workforce.World Economic Forum6 Roland Berger,Microsoft,NVIDIA|Accelerate,innovate,empowerSource:Roland BergerPotential impact of AIAI use cases(selection)AI value-addService&customer mgmt.HighLow Au

23、tomize customer service(chatbots)Compile customer reviews database to identify priority areas for improvement Compile FAQs Improvement in response time FTE cost savings Streamline customer data acquisitionMarketing&salesHighLow Develop website content and sales materials AI-powered CRM to personaliz

24、e sales strategy&process Fast and in-depth price analytics Less human effort in training and sales material development Higher sales conversion ratesSupply chain&logisticsHighLow Shipping routes and delivery times analysis Streamline AVL mgmt.monitoring and recording suppliers performance Fleet and

25、time mgmt.optimization Reduced delivery CO2 impact Accelerated risk mitigationProduction managementHighLow Vision-based automated quality inspection Equipment condition monitoring&predictive maintenance Troubleshooting and PLC programming Increased throughout&reduced lead time Reduction of downtimes

26、&increased utilization OEE increaseProcurementHighLow Data-driven category strategy development Tender document creation Contract authoring and review Maximized data transparency Increased number of market interactions Enhanced value contribution(savings,risk mitigation,innovation,sustainability,com

27、pliance)DesignHighLow Predictive 3D design generation for parts and process optimization Screen design database to benchmark market solutions Automated and error-free design from standardized frameworksResearch&developmentHighLow Scan patent database to provide compliance check Generate technical sp

28、ecs based on RFQ and client standards BOM optimization,screening cost-reduction solutions R&D budget cost savings Faster development cycles&reduced TTM Higher performance with optimized BOMOptimized and automated SOP and inventory managementA AI encompasses a range of technologies,with GenAI emergin

29、g as the latest addition to the family7 Roland Berger,Microsoft,NVIDIA|Accelerate,innovate,empowerWhy is this the case?Besides speed,many other major obstacles stand in the way of commercial AI adoption.Primary among these are proving business value,missing infrastructure to support the case and a s

30、hortage of skilled AI talent.C Despite these challenges,companies cannot afford not to act.They need to embrace intelligent solutions for next-generation manufacturing and operations more than ever,as they can play a key role in navigating todays economic,environmental and societal challenges.Howeve

31、r,increasing the adoption rate of AI in operations will require more than merely launching increasing numbers of pilots.Instead,to take full advantage of the value from the multitude of use cases,it is time to take stock and define a flight path.This is a holistic approach that determines the right

32、solutions to pursue,leverages the right enablers at the right time and as part of an integrated strategy,and ultimately creates tangible impact quickly,but in a streamlined,focused,pragmatic fashion.B GenAI will lead to significant performance improvements and will boost results along all typical pr

33、ocurement goals Estimated impact of GenAI(by leveraging LLMs)on procurement goals(extract)Source:Roland BergerProductivityimprovementEfficiencyincreaseAI boostSustainability&complianceInnovationSpendreductionRisk mitigationProcurement compliance100%Lower CO2 emissions30%More innovativesupplier colla

34、borations50%Reduced excess inventory&stockouts30%Faster procurement processes4 xImproved FTEefficiency50%More supplier interactions60%Less supply chain disruption40%Lower market price risks30%Higher savings(e.g.,6%instead of 4%)50%8 Roland Berger,Microsoft,NVIDIA|Accelerate,innovate,empowerA WAY FOR

35、WARDTo optimize the AI in the operations flight path,business must consider three key strategic questions:What are the value levers?Before thinking about technology,identify pressing problems and value pockets.Without considering the implications for AI,clarify the strategic optimization opportuniti

36、es in operations.How do I leverage the technology?As a second step,mirror the identified opportunities with technological capabilities.What technology might bring value to my operations?Which use cases are worth pursuing,not as an end in themselves,but to strategically improve my processes?Which tec

37、hnologies are at hand and can be leveraged,and which need to be built up?And which skills and capabilities are required and will need to be developed?Source:StatistaChallenges in measuring and proving the business value of the AI solutionLack of technological infrastructure to support AIShortage of

38、skilled AI talentLack of clean dataLack of trust toward AI-based decisions45%40%35%30%25%20%15%10%5%0%C Top AI adoption challenges:Proving business value and a lack of technological infrastructure22%42%38%32%24%9 Roland Berger,Microsoft,NVIDIA|Accelerate,innovate,empowerWhat is my roadmap?Finally,pl

39、ayers need to develop a roadmap,defining the steps involved in building appropriate capabilities and integrating them in functional teams,quantifying benefits and creating a tangible pathway toward value creation.The final question is,how?In the following chapters,we provide practical examples of us

40、e cases,key success factors and details of how Roland Berger and Microsoft can support companies in navigating the AI in the operations flight path.Key success factors for AI in operations Our analysis shows that industrial digital leaders have seven AI adoption principles in common,from clear roadm

41、aps to accessible data and empowered workers Over the past five years,the MIMA application rounds helped to build a comprehensive picture of what digitalization can achieve in operations.This has enabled us to identify seven key factors common to successful adopters of digital solutions,and in parti

42、cular AI in operations.All should be considered by companies when developing their flight path toward impactful digitalization.In the following pages,we outline these key success factors and present tangible,real-world examples of each from among MIMA entrants.2.1/Develop a clear AI roadmap with exp

43、licit objectives Outline the change process,match capabilities with targets and ensure flexibility to maximize value-addThe vast majority of companies lack a clear roadmap from the endless possibilities and use cases AI offers operations to tangible applications for specific day-to-day needs.To addr

44、ess this,successful AI-adopting companies develop a structured plan for AI applications within the organization,based around three steps:D2 Correlating data from over 100 process steps,DXQanalyze actively supports the operator in reducing the amount of rework required.Numerous OEMs have successfully

45、 deployed the software and,as a result,have benefited from considerable savings.Gerhard Alonso Garcia,VP MES&Controls,Drr Systems AG10 Roland Berger,Microsoft,NVIDIA|Accelerate,innovate,empowerIdentify key challenges that can be solved using AI-enabled solutionsDownstream quality control of car body

46、 painting lines takes up to two weeks between error occurrence and mitigation,a long period during which the error is repeated.Rework rates can be as high as 20%,a significant cost for OEMs.Drr,a leading global mechanical and plant engineering firm,tackled this key issue:AI-driven software has been

47、pivotal in positioning OEMs at the forefront of their industries,hence Drr used AI to turn painting line data into quality insights in almost real time.The in-line,edge-device-based analysis system DXQanalyze detects and learns from anomalies in the painting process,to predict problems before they o

48、ccur.Determine which technologies are available,and what domain expertise can be leveraged to generate AI value-add Successful leveraging of AI requires coupling the algorithmic with the appropriate domain expertise.Malvern Panalytical,a manufacturer and supplier of laboratory instruments,coupled it

49、s immense domain expertise in common analytics systems for agricultural with new AI capabilities,building a near-infrared handheld sensor with Microsoft Azurebased platform and algorithmic capabilities.The resulting solution provides recommendations on appropriate fertilization to farmers in less th

50、an a day.This increases farming efficiency(by 20%30%for bananas,e.g.)while decreasing the environmental burden of overfertilization.Danieli Automation,part of the Danieli Group,a plant supplier,can draw on extensive knowledge in steelmaking processes,both from the wider equipment manufacturing group

51、 and from the groups specialist steelmaker subsidiary Acciaierie Bertoli Safau.For its Q3 platform,the company combined its domain expertise and analytics capabilities to create an end-to-end quality assurance solution for the complex steelmaking process,reducing physical sampling efforts by 20%.Cre

52、ate a game plan for an AI-enabled operations future,with executive sponsorship and a clear organizational change processDigital leaders create holistic game plans considering the entire organization,as well as structured prioritization of use cases along systematic criteria.Faced with the digital tr

53、ansformation of 200 plants worldwide,the food company Danone set up its holistic Digital Manufacturing Acceleration program.It consists of three action streams:blue,which includes activities supporting local management to identify and prioritize use cases in a structured way,for example by providing

54、 a catalog with established solutions from the Danone group;yellow,which is fully dedicated to upskilling on change management and digital competencies;and red,which comprises technical standardization and security topics.The plan has already realized annual savings of EUR 55 million across 71 facto

55、ries.KEY TAKEAWAYS Identify key problems and match the right expertise and technology to tackle them.Avoid embarking on siloed pilots and instead follow a holistic AI roadmap aligned with a holistic,executive-led corporate strategy Focus on maximum impact,flexibility to adapt to technological change

56、s and required organizational change11 Roland Berger,Microsoft,NVIDIA|Accelerate,innovate,empower2.2/Determine the technical foundation&ensure data accessibility.Develop a flexible,broad-based IT foundation and make data sufficiently and securely accessibleTo ensure the existing system landscape has

57、 the potential to meet the roadmap objectives,AI-in-operations leaders focus on two main areas:Legacy systems and IT requirementsBeyond algorithms,a digital factory requires plenty of hardware to ensure optimal performance of intelligent manufacturing systems.BMW Group,the carmaker,developed its Edg

58、e Ecosystem platform to maintain this hardware.Countless edge computing devices ensure swift and efficient computing in proximity to the shop floor and its sensors.In BMWs iFACTORY concept,these devices are centrally managed over the cloud,enabling consolidation of shared data and the rapid swapping

59、 out of malfunctioning devices by shifting software onto another device,minimizing downtime.As a result,asset management effort is reduced by 80%,the company says.Use case data requirements and data accessibility/securityTo maximize internal data accessibility,the carmaker Mercedes-Benz created its

60、MO360 Data Platform.It decouples data storage and usage,with all contained operations data stored in a Source:StatistaD AI adoption perceptions Only 20%of business leaders believe their organizations often apply AI solutionsExtent to which commercial leaders feel their organizations should be using

61、ML or GenAI in 2023Extent to which commercial leaders feel their organizations are using ML or GenAI in 2023 Machine learning(ML)Generative AI (GenAI)50%25%20%20%65%0%0%55%RarelyRarelyOftenOften12 Roland Berger,Microsoft,NVIDIA|Accelerate,innovate,empowerstructured manner and accessible via a self-s

62、ervice marketplace.This allows swift use-case realization for users across all functions,from shop floor to engineering.The high level of accessibility leads to tangible results,with around 80 use cases already in use or in the pipeline two years since launch.A 20%operations efficiency increase is e

63、xpected by 2025.Externally,digital twins are a well-known principle to store all data related to a product in a structured manner.The drive system manufacturer Wittenstein has a textbook example,shipping all products with a digital twin that is compliant with Asset Administration Shell and IEC 61406

64、 Identification Link,and either accessible by data matrix code or integrated into smart products.Data is tracked throughout the entire life cycle,from development to after-sales service,enabling optimized product design,commissioning,field performance and sustainability.Meanwhile,to maximize trust i

65、n data,German multinational Siemens extended the digital twin concept to a secure,blockchain-based genealogy tree with its Trusted Traceability solution.The blockchain is fed with,for example,processing,manufacturing and component information along its life cycle,preventing counterfeiting of or tamp

66、ering with products such as critical circuit boards.KEY TAKEAWAYS Create a technical foundation that supports a broad spectrum of applications,avoiding siloed solutions.Ensure it is flexible and future-proof.Microsoft Fabric is a good example of a one-stop-shop solution,allowing integrated data mana

67、gement,processing,analytics and reporting Make structured,standardized data(and interfaces)accessible across the business these remove fear of contact,allow for a fail fast approach and accelerate use case results2.3/Embrace modular platforms.Modular platforms allow AI solutions to be flexible,scala

68、ble and easily upgradeableThe potential of AI-driven solutions lies in their ability to connect the dots,analyze large amounts of data and support good decision-making based on broad insights.A holistic approach that overcomes solution silos is therefore crucial.Austrian production equipment manufac

69、turer Andritz faced this challenge,with many of its clients pulp factories characterized by process islands and silo systems.So,the company systematically mapped the typical activities of plant staff and assessed which could be enhanced using AI.It then consolidated solutions in its Metris platform.

70、This enabled overarching solutions,for example automated anomaly management or an operational risk score assessment.Almost 40 plants worldwide adopted the systems,with an 18%production increase recorded in some factories.KEY TAKEAWAYS A modular platform approach to AI solutions simplifies scaling,ad

71、aptation and enhancement of solutions with upcoming technologies,whether in-house or third-party.Microsofts Azure Marketplace provides a valuable resource and partner network to flexibly use AI solutions,and is simple to procure and deploy in an Azure environment 13 Roland Berger,Microsoft,NVIDIA|Ac

72、celerate,innovate,empower2.4/Leverage off-the-shelf technology components and take a fail fast approachReadily available solutions are tried and tested,cut development times and reduce risk A lack of technological infrastructure and skilled AI talent rank in the top three industrial AI adoption chal

73、lenges.However,low code platforms,such as Microsofts Power Platform and Power Apps,and an abundance of available libraries and tool components now allow solutions to be created swiftly and simply using off-the-shelf components without significant investments.These tools enabled Swedish battery manuf

74、acturer Northvolt to solve the major challenge posed by small particles,which risk battery performance and employee health,in less than two weeks.The TechClean solution combines,for example,image recognition and AI-enabled analytics to provide operators with text-based support for the handling of pa

75、rticle traps.Despite the short lead time and use of standard components such as Power Platform,the company was able to realize savings of around 5,200 hours per month.KEY TAKEAWAYS Leverage off-the-shelf technology components to accelerate development,improvement and deployment of smart solutions,wh

76、ile maintaining solution reliability and keeping effort and cost to a minimum Use such solutions to enable a fail fast approach,generating hands-on AI application experience for the organization with minimal economic risk.Complexity is also reduced,further limiting risk2.5/Democratize digitalization

77、Ensure employee buy-in,promote innovation and facilitate the spread of successful solutions Acceptance of novel AI-enabled tools across an organization is crucial to their success,as is engaging people to help shape the digitalization of operations.Mercedes-Benzs MO360 platform not only gives users

78、easy self-service access to data,but also encourages all levels of the organization to develop and prototype their own cases,for example via ideation pitches and dedicated communications.Northvolts rapidly realized TechClean solution also demonstrates how seemingly small-scope solutions that target

79、a specific on-the-job problem can,when combined,outperform lighthouse use cases,which often have limited traction in day-to-day operationsKEY TAKEAWAYS Educate and encourage users at all levels to reflect on AI usage potentials in their respective departments or functions E Foster a spirit of AI-bas

80、ed innovation,encouraging the rapid development of seemingly small,pragmatic cases,such as off-the-shelf components,for example by leveraging citizen development tools such as Microsoft Power Platform14 Roland Berger,Microsoft,NVIDIA|Accelerate,innovate,empower2.6/Empower the workforceTarget AI solu

81、tions on tasks where they can best support and free up human workers AI-enabled algorithmic solutions are great at processing high volumes of data in-line to provide decision support or guidance in real time.However,some processes still require the intuition and skill of manual workers.AI-in-operati

82、ons leaders therefore:Leverage human and AI strengths to maximize operational effectivenessSiemens Gamesa has combined human and AI strengths in wind turbine blade production.The process requires precise manual placement of 1,400 delicate fiberglass mats into a 108-meter-long mold with millimeter pr

83、ecision.After the resin is cured,error correction is complicated and costly.However,the company developed an algorithmic AI solution based on overhead cameras that uses lasers to guide workers as they place the mats.Its success meant the solution broke even in just 2.5 years.EAIs potential:Improveme

84、nt potential estimated by R&D experts Conventional AI and GenAI solutions are expected to deliver significant boosts in business efficienciesSource:Roland BergerPerformanceimprovementEfficiencyincreaseAI boostSustainability&complianceInnovationRisk mitigationProduct cost optimizationIncreased patent

85、 compliance35%More patent applications60%Higher benchmarking accuracy50%First time right:faster requirement checks&testing cycles100%Lower CO2 emissions25%Improved FTEefficiencyReduced product development time,faster development cycles35%60%Higher number of simulation cycles,enhanced quality55%Lower

86、 prototyping costs40%R&D budget optimization&cost savings25%Better functionality and manufacturability60%15 Roland Berger,Microsoft,NVIDIA|Accelerate,innovate,empowerMake the best use of declining workforce numbersGerman pump manufacturer Wilo provides workers with an assistance system that adapts a

87、ssembly instructions to the observed skill level.It even includes a gamification feature,which motivates employees,while ensuring all employees are provided with the optimal level of support.Globally,the company attributes a more than 20%productivity increase,60%reduced customer claims and a EUR 5 m

88、illion reduction in non-conformance costs potential to the workforce assistant,as well as noting an invaluable boost in motivation at a time of labor shortages.The machine tool producer Trumpf optimizes the use of scarce tooling machine operators by geographically detaching machine and operator.Its

89、remote monitoring and control service allows machine specialists to oversee many geographically distributed machines.While 80%of Trumpfs clients report problems finding adequate workforce,Trumpf has increased productivity and gained valuable machine data for analytics purposes.Use AI to eliminate te

90、dious,repetitive tasks,freeing up employees for more creative workBuilding inspection is a repetitive process,requiring regular photographic documentation of the site.Construction company Goldbeck has automated the process using a dog-shaped robot thats programmed to conduct regular photo-documentat

91、ion walks of a building site.The system preventively points things out that I wouldnt have seen otherwise,says Maximilian Schuetz,Goldbecks Director of Building Information Modeling.It also allows workers to spend more time managing the site.KEY TAKEAWAYS Identify the respective strengths of AI and

92、the human workforce Exploit AI and digital solutions to enable skilled workforce to focus on value-add activities Diffuse GenAI-powered solutions within the organization to eliminate usage concerns and harness direct productivity benefits,for example through Microsoft Copilot technology,such as the

93、Dynamics 365 or GitHub Copilot 2.7/Exploit AIs potential to disrupt value creationUse AI-generated insights to improve offering,maintain installed base and create new productsBeyond efficiencies in existing products and processes,AI has the potential to entirely disrupt traditional ways of generatin

94、g revenue and ownership.AI-generated insights into the(digitalized)installed base of a product can enable manufacturers to improve their offering or better understand the reliability or maintenance profiles of the installed base.In We have already increased the productivity of our customers by 50%.D

95、r.-Ing.Philipp Humbeck,R&D Manager at Trumpf16 Roland Berger,Microsoft,NVIDIA|Accelerate,innovate,empowerturn,this can leverage additional non-technological value pockets,for example by pooling insurance policies in cases where the installed base remains largely the property of the manufacturer.For

96、example,Andritz leverages data from its clients digitalized pulp factories to sell production guarantees instead of bare products.So,the company can participate in the efficiency gains realized across its deployed product fleet while simultaneously reducing client downtime risk.Trumpf has developed

97、a similar machine-level system.It offers its remotely controlled machines on a pay-per-part basis,leveraging the economies of scale of its highly efficient centralized machine control operations to cut client downtimes and machine underutilization.Meanwhile,Danish heating and cooling equipment manuf

98、acturer Danfoss entirely turned around its product offering from physical HVAC components to climate solutions.Its fully connected Alsense platform increases efficiency across the board and prevents unforeseen maintenance through AI-enhanced analytics.Similarly,Boschs Blaichach plant showcases how a

99、 solution originally geared toward an internal challenge can be turned into a new product.Building upon image recognition algorithms,the engineering and tech multinational developed a solution to rapidly identify required spare parts from its extensive parts repository,reducing search time for spare

100、 parts by 80%.After roll-out to 10 Bosch plants already and digitization of more than 180,000 parts,the solution is now being implemented at first third-party customers as well,enhancing value creation within existing processes while adding a new offering.KEY TAKEAWAYS Harness AI-generated insights

101、to offer new products or improve the performance of the installed base Spin off successful AI solutions into dedicated products to create entirely new business opportunities,for example by leveraging AI agents/copilots for external customer use to improve clients productivityOutlook:The future impli

102、cations of AI The showcased applications demonstrate that companies are already benefiting from using AI to enhance their operations.However,they have only scratched the surface of its full potential.Companies are now actively exploring ways to accelerate its integration,foster innovation within the

103、ir organizations and empower their workforce with AI-based tools.In parallel,the AI technology underlying the realized and potential applications is evolving rapidly:development of AI models is expected to accelerate exponentially due to ongoing training on more and more data parameters to increase

104、fit and quality of AI output.Models are becoming more versatile and capable of flexibly handling various data types and input such as text,audio/voices,video and source code,further enhancing both applicability and usability of AI tools.As we see with recent skyrocketing advancements in generative A

105、I,AI technology will become more self-sufficient and independent from predetermined structures in data.It will learn data patterns on its own,provide proactive suggestions for improvements and 317 Roland Berger,Microsoft,NVIDIA|Accelerate,innovate,empowerultimately develop solution pathways to solve

106、 given challenges independent of human involvement.As a result,the technology will seamlessly integrate into all content-generating activities,in both private and commercial contexts.This will have three key implications for individuals and companies:1 It is essential for individuals and companies t

107、o embrace and utilize AI to remain competitive in the medium term through efficiency gains and productivity.This will ultimately improve profit margins on a company level and contribute positively to the overall GDP.2 To fully implement the technology in a far-reaching way,companies need to bring ne

108、w skills and capabilities into their organizations.AI and especially GenAI require new disciplines to be taught to employees and integrated into functional teams,such as data science,prompt engineering or data management.3 With the increasing autonomy of AI,ethical considerations and compliance beco

109、me crucial topics.Companies will need to prioritize these factors and ensure that their use of the technology aligns with ethical standards and regulatory requirements.NVIDIA:An expert perspective on the future of AI in operationsThe integration of AI in the manufacturing industry offers companies g

110、reat potential to increase their competitiveness,says Dr.Timo Kistner,EMEA Manufacturing Lead at full-stack AI company NVIDIA.But what should businesses expect from AI?In this section,he outlines some key trends in AI use cases,highlights the growing importance of generative AI in data acquisition a

111、nd offers tips for successful implementation of AI tools.How will new AI tools disrupt business models and value chains in manufacturing?Artificial intelligence will influence the entire value chain of manufacturing companies,from development to supply chain optimization and production,as well as in

112、 marketing,service and administration.In the production environment,AI is already supporting the optimization of quality,production volume and plant availability,thereby making a significant contribution to improving overall efficiency in manufacturing.The advancement in generative AI in particular

113、the use of large language models now makes it possible to gain insights from a wide variety of data silos that were previously hard to combine and analyze.This provides early-warning information about the availability of systems and necessary maintenance cycles,as well as insights into optimization

114、potential that can only be tapped on the basis of this technology in the first place.What are the key trends and technological advancements that will shape AI use cases in the short-term?First of all,AI is being brought ever further into the virtual world,changing the way we work with the technology

115、.For example,today we develop new AI products in the virtual space,not on the factory floor.Then we move to the physical world to try them out.This is much more efficient as there is no need to go back and forth while developing products.Secondly,we are now using physical features(factories,products

116、,etc.)to train AI in the virtual world so that it can better map the physical world.418 Roland Berger,Microsoft,NVIDIA|Accelerate,innovate,empowerData is the key to AI.Yet companies complain of being constantly short of good data.What developments do you envision in data acquisition?A key issue for

117、the efficient use of AI is certainly the availability and quality of data for training AI applications.Today,generative AI helps create synthetic data that can be used to train high-quality AI models using only a small number of,for instance,reference images,helping to fill the data gap.For example,

118、if there are insufficient pictures for use in a quality-control application,generative AI can create synthetic data,be it from pictures or natural language.At the same time,generative AI can help to scale corresponding models as synthetic data can feed a broad number of use cases.The technology also

119、 enables smaller and medium-sized companies with comparatively little expertise to leverage AIs potential.How can companies best plan for and implement AI?The use of AI for its own sake is not effective;it requires a clear strategy(as outlined in this reports key success factors).A focused approach

120、to the technology,an analysis of the potential use cases,as well as the return on investment,are fundamental requirements for the successful implementation of use cases.Artificial intelligence offers great potential for optimizing business processes,but at the same time it also requires a transforma

121、tion of the company to ensure applications are used efficiently.The transformation of processes,and in particular the involvement of all employees as part of the transformation process,is a core element of the successful implementation of an AI strategy and must be driven forward by management.How w

122、e can support AI in Operations Roland Berger and Microsoft have respectively developed a multitude of tools to leverage the benefits of AI for companies While the key success factors of AI leaders in operations are clear,it is less clear how companies in the early stages of AI adoption can emulate t

123、hem.Indeed,preparing for and embarking on the journey to successful AI in operations can seem a little overwhelming.To support such companies,Roland Bergers Industrials,Operations and Digital competence centers joined forces to develop the comprehensive rAIse framework and several other proprietary

124、tools that add significant value to its consulting services.As a leading developer of enterprise technology solutions,Microsoft also has a stable of cutting-edge solutions harnessing the power of AI.Roland Berger:rAIsing the barThe rAIse framework is centered on three core pillars for building a hol

125、istic pathway to AI adoption people,processes&culture;tech&execution;and strategy&value proposition.Each incorporates several value creation indicators,from strategic aspiration and organizational setup to technical blueprint and operationalization.FAt the very start of the rAIse process,Roland Berg

126、ers global experts use our proprietary AI Readiness Assessment to identify existing and missing capabilities within each indicator.Required next steps can then be tailored toward targeted recommendations.519 Roland Berger,Microsoft,NVIDIA|Accelerate,innovate,empowerDepending on individual organizati

127、ons capabilities and needs,rAIses modular approach can then be customized to focus on either strategic roadmapping or distinct components.For example,the latter might include an AI North Star definition or the design of a future,AI-boosted operating model catering to elevating RBs services.To comple

128、ment rAIse and allow an immediate impact and swift,tangible results,Roland Berger can draw from an extensive and constantly growing use case library.This covers all functions and the entire operations value chain,from R&D to manufacturing to finance and accounting.GFurthermore,Roland Berger can leve

129、rage its data analytics and digitalization powerhouse RB N3XT.This consists of a global team of dedicated specialists merging data science and data engineering with strong consulting capabilities.Extending well beyond strategy building,N3XT enables Roland Berger to rapidly develop AI-empowered proto

130、types and tools or tailor existing,proprietary solutions that bring its consulting services to a new level.For example,RBs proprietary GenAI Platform uses large-language-model-based AI agents to securely process client data and rapidly generate the best possible answers to pressing business challeng

131、es.HFThe rAIse framework:Our approach encompasses 10 value creation indicators Vision and value creationTech FoundationPeople&Organization EnablersAI Vision&AspirationsAI Development&OperationalizationStrategy&value propositionTech&executionPeople,processes&cultureOperating ModelCompute&TechInfrastr

132、uctureGovernance&PoliciesData FoundationHuman CapitalTechnical BlueprintPartnership&Sourcing ModelAI Value Creators 10 9 8 7 1 2 3 4 5 620 Roland Berger,Microsoft,NVIDIA|Accelerate,innovate,empowerStrategy&corporate developmentR&D&product designSoftware development10+functions450+use casesKnowledge

133、managementITFinance&accountingMarketing&salesCustomer operationsService deliveryManufacturingSupply chainProcurementSource:Roland Berger G Use case library:Example applications are available from across the value chain21 Roland Berger,Microsoft,NVIDIA|Accelerate,innovate,empowerHGenAI platform:The a

134、pproach generates answers to key business challengesSource:Roland Berger Understand your business questions and are trained to interact with your data Learn the patterns and make-up of your unstructured data and can generate new insights Can be completely tailored to fit your situation and business

135、process State-of-the-art large language models are the engine that power the AI agentsGenAI platformWeb appBusiness questions Secure storage of your data/documents on your preferred infrastructure (e.g.,SharePoint)How to make our products more competitive?How to significantly reduce third-party spen

136、d?How to get to better decisions?How to improve product profitability?How to boost sustainability?How to increase internal knowledge access?LLMs (GPT,Gemini,Mistral,)AI agentsCompany data22 Roland Berger,Microsoft,NVIDIA|Accelerate,innovate,empowerFinally,tools such as the proprietary RB Knowledge H

137、ub for GenAI-based leveraging of existing data and expertise enable Roland Berger to apply the power of AI to project execution,ensuring swift,data-backed insight generation for our clients.Microsoft:A trusted Copilot The beginning of an AI journey can be overwhelming.But a well-structured plan buil

138、t around Microsoft tools can ease the transition.First of all,successful integration of AI requires a solid foundation.Companies should begin by migrating applications,infrastructure and the data estate,including SQL,governance and analytics.This step is crucial for a smooth transition to AI-powered

139、 operations.For companies new to AI,Microsoft then recommends exploring Copilot Microsofts generative AI work companion within the Microsoft suite.Copilot solutions are easy to implement,ensuring users can leverage the benefits right away.For example,Copilot can be used to summarize meetings or writ

140、e emails.Once familiar with the technology,the more advanced Dynamics 365 Copilot,Copilot in Power Platform or GitHub Copilot can be used to enhance operations.Dynamics 365 Copilot automates and optimizes business processes within Dynamics 365,improving sales through intelligent suggestions and data

141、 enrichment,for example.GitHub Copilot empowers developers with AI-assisted coding capabilities,making software development faster and more intuitive.These tools are designed to streamline workflows,boost productivity and provide actionable insights through AI-driven functionalities.IIf there is a u

142、se case in mind,Microsofts partners can assist in developing tailored AI solutions,whether its creating an intelligent app or designing a custom AI assistant.For a faster time-to-value,companies can explore AI solutions available in the Azure Marketplace.These off-the-shelf tools can be swiftly impl

143、emented into a companys own intelligent IPlug and play:Many of Microsofts Copilot solutions are ready to use How you can get started todayTry Microsofts Copilots in actionEnsure an AI-ready data foundationGet Copilot-ready with Microsoft 365Build your own AI-powered Copilot23 Roland Berger,Microsoft

144、,NVIDIA|Accelerate,innovate,empowersolutions,as demonstrated in the Northvolt use case,and can be easily deployed in the Azure environment,providing immediate benefits.In addition,Microsoft provides a wide range of support mechanisms to facilitate AI journeys,with access to extensive AI capabilities

145、.Co-development and co-engineering partnerships allow for the creation of AI solutions specific to individual business requirements,for example,while advisory services provide expert guidance to help strategize and implement tailored AI solutions.Microsoft also offers implementation support,assistin

146、g in deploying and integrating AI technologies into a companys operations.Co-investments help to accelerate AI adoption.Additionally,Microsoft has an extensive network of partners for additional support and resources.JFor more information about how Roland Berger and Microsoft respectively leverage A

147、I tools and expertise,and how AI in operations can transform your business,please contact one of Roland Bergers or Microsofts respective experts.We look forward to hearing from you and supporting you on your journey to AI-powered success.J Assisted development:Microsoft products offer support throug

148、hout the AI transformation journeyTech stackMicrosoft Support MechanismsCollaborate,innovate,secure,optimize and scale solutions effectivelyAdvisoryImplementationCommercialCo-developmentPartner networkSource:Roland Berger24 Roland Berger,Microsoft,NVIDIA|Accelerate,innovate,empowerAUTHORS ROLAND BER

149、GERGERMANYJochen GleisbergSenior Partner Operations Pieter NiehuesPartner Operations and AI expert Julia DuwePartner Operations and AI Sven MarlinghausSenior Advisor OLasse AdlerPrincipal OFelix MohrschladtSenior Consultant IUSAMichelle Drew RodriguezPartner Industrials ITALYMaria MikhaylenkoSenior

150、Partner Global Managing Director and Digitalization EAUTHORS MICROSOFTMatt Walsh Managing Director EMEA Manufacturing&Supply Chain lead Paul Rhrs Director Industry Advisor Manufacturing&Mobility Ralf Schnfeld Principal Industry Advisor Healthcare&Life SciencesNico Hartmann Manufacturing Industry Lea

151、d GermanyAUTHORS NVIDIATimo Kistner EMEA Industry Lead AI for Manufacturing&IndustrialNicolas Savides Hyperscalers Alliances Manager EMEAThis whitepaper was created with the friendly support of Hasmeet Kaur,former Partner and Global Managing Director at Roland Berger,leading Roland Bergers Innovatio

152、n and AI initiatives.Furthermore,we thank Tahir Simsek,Microsoft Industry Advisory,for his contribution to the whitepaper.CONTACTS ROLAND BERGERBRAZILCristiano DPartner Industrials/OperationsSo PauloGERMANYOliver KPartner OperationsStuttgartThomas KPartner IndustrialsHamburgAUSTRIAGundula PPartner O

153、perationsViennaFRANCEMichel JPartner Operations/IndustrialsParisMagali TPartner OperationsParisEric KPartner IndustrialsParisDUBAIGabriella BPrincipal OperationsDubaiBELGIUMAxel BPartner ServicesBrusselsGrgoire TPartner Transaction&Investor ServicesBrusselsITALYAlfredo APartner Services/Transaction&

154、Investor ServicesMilanSPAINPol BPartner Industrials/Health&ConsumerMadridUKHouda OPrincipal OperationsLondonThis whitepaper is for informational purposes only and is not offered as professional advice for any specific matter.Professional advice should always be sought before taking any action or ref

155、raining from taking any action based on this whitepaper.Roland Berger group of companies(Roland Berger),Microsoft group of companies(Microsoft),and the editors and the contributing authors do not assume any responsibility for the completeness and accuracy of the information contained therein and exp

156、ressly disclaim any and all liability to any person in respect of the consequences of anything done or permitted to be done or omitted to be done wholly or partly in reliance upon the whole or any part of the whitepaper.The whitepaper may contain links to external websites,and external websites may

157、link to the whitepaper.Roland Berger and Microsoft are not responsible for the content or operation of any such external sites and disclaim all liability,howsoever occurring,in respect of the content or operation of any such external websites.SWEDENHauke BDirector Operations/IndustrialsStockholmJAPA

158、NMasashi OPartner OperationsTokyoCHINALiang QPartner Operations/RPTShanghaiMALAYSIAChris OPrincipal Operations/Health&ConsumerKuala Lumpur 2024 ROLAND BERGER GMBH.ALL RIGHTS RESERVED.09.2024ROLANDBERGER.COM24_2423_REPROLAND BERGER is one of the worlds leading strategy consultancies with a wide-rangi

159、ng service portfolio for all relevant industries and business functions.Founded in 1967,Roland Berger is headquartered in Munich.Renowned for its expertise in transformation,innovation across all industries and performance improvement,the consultancy has set itself the goal of embedding sustainabili

160、ty in all its projects.Roland Berger revenues stood at more than 1 billion euros in 2023.Founded in 1975,MICROSOFT(Nasdaq:MSFT)is one of the worlds leading software services,and solutions companies.It helps people and businesses realize their full potential.The technology company creates platforms a

161、nd tools powered by AI to deliver innovative solutions that meet the evolving needs of its customers.Microsoft is committed to making AI available broadly and doing so responsibly,with a mission to empower every person and every organization on the planet to achieve more.The MICROSOFT INTELLIGENT MA

162、NUFACTURING AWARD(MIMA),jointly organized by Microsoft and Roland Berger,celebrates thought leaders who are shaping the future of industry with their ideas and solutions.Awarded annually in six categories,it honors the most promising and innovative digital best practices from the industrial and operations sectors,e.g.,production,purchasing,supply chain,engineering and after-sales.Stay tuned and apply with your pioneering use case for the MIMA 2025 on !PublisherRoland Berger GmbHSederanger 180538 MunichGermany+49 89 9230-0