《WEF&BCG：2025工業運營前沿技術：AI智能體 (AI Agents) 的崛起白皮書（英文版）（26頁）.pdf》由會員分享，可在線閱讀，更多相關《WEF&BCG：2025工業運營前沿技術：AI智能體 (AI Agents) 的崛起白皮書（英文版）（26頁）.pdf（26頁珍藏版）》請在三個皮匠報告上搜索。

1、Frontier Technologies in Industrial Operations:The Rise of Artificial Intelligence AgentsW H I T E P A P E RJ A N U A R Y 2 0 2 5In collaboration with Boston Consulting GroupTransformation of Industries in the Age of AIImages:Getty ImagesDisclaimer This document is published by the World Economic Fo

2、rum as a contribution to a project,insight area or interaction.The findings,interpretations and conclusions expressed herein are a result of a collaborative process facilitated and endorsed by the World Economic Forum but whose results do not necessarily represent the views of the World Economic For

3、um,nor the entirety of its Members,Partners or other stakeholders.2025 World Economic Forum.All rights reserved.No part of this publication may be reproduced or transmitted in any form or by any means,including photocopying and recording,or by any information storage and retrieval system.ContentsRea

4、ding guide 3Foreword 4Executive summary 5Introduction 61 The next leap:reinventingindustrial operations through 7 frontiertechnologies1.1 Entering the next frontier:the path towards self-control 81.2 Redefining the role of humans:from operators 8 to AI-enabled orchestrators2 AI agents fuelling the t

5、ransformation ofoperations 102.1 Virtual AI paving the way for autonomous systems 122.2 Embodied AI igniting a new era in robotics 153 Strategic imperatives for industrial operations transformation 183.1 Paving the way for successful use of AI agents 18 in industrial operations3.2 Staying at the for

6、efront of AI agent innovations 193.3 Building the foundations:organizational and technological 19Conclusion 21Contributors 22Endnotes 25Frontier Technologies in Industrial Operations2Reading guideThe World Economic Forums AI Transformation of Industries initiative seeks to catalyse responsible indus

7、try transformation by exploring the strategic implications,opportunities and challenges of promoting artificial intelligence(AI)-driven innovation across business and operating models.This white paper series explores the transformative role of AI across industries.It provides insights through both b

8、road analyses and in-depth explorations of industry-specific and regional deep dives.The series includes:As AI continues to evolve at an unprecedented pace,each paper in this series captures a unique perspective on AI including a detailed snapshot of the landscape at the time of writing.Recognizing

9、that ongoing shifts and advancements are already in motion,the aim is to continuously deepen and update the understanding of AIs implications and applications through collaboration with the community of World Economic Forum partners and stakeholders engaged in AI strategy and implementation across o

10、rganizations.Together,these papers offer a comprehensive viewof AIs current development and adoption,aswell as a view of its future potential impact.Each paper can be read stand-alone or alongside the others,with common themes emerging acrossindustries.Frontier Technologies in Industrial Operations:

11、The Rise of Artificial Intelligence AgentsW H I T E P A P E RJ A N U A R Y 2 0 2 5In collaboration with Boston Consulting GroupTransformation of Industries in the Age of AIArtificial Intelligence inFinancial ServicesW H I T E P A P E RJ A N U A R Y 2 0 2 5In collaboration withAccentureTransformation

12、 of Industries in the Age of AIAI Governance AllianceThe Future of AI-Enabled Health:Leading the WayW H I T E P A P E RJ A N U A R Y 2 0 2 5In collaboration withBoston Consulting GroupTransformation of Industries in the Age of AILeveraging Generative AI for Job Augmentation andWorkforce Productivity

13、:Scenarios,Case Studiesand a Framework for ActionI N S I G H T R E P O R TN O V E M B E R 2 0 2 4In collaboration with PwCArtificial Intelligences Energy Paradox:Balancing Challenges and OpportunitiesW H I T E P A P E RJ A N U A R Y 2 0 2 5In collaboration withAccentureTransformation of Industries i

14、n the Age of AIAI Governance AllianceIn collaboration with the GlobalCyber Security Capacity Centre,University of OxfordArtificial Intelligence andCybersecurity:Balancing Risks andRewardsW H I T E P A P E RJ A N U A R Y 2 0 2 5Transformation of Industries in the Age of AIAI Governance AllianceAI in

15、Action:Beyond Experimentation to Transform IndustryF L A G S H I P W H I T E P A P E R S E R I E SJ A N U A R Y 2 0 2 5In collaboration withAccentureAI Governance AllianceTransformation of Industries in the Age of AIIntelligent Transport,Greener Future:AI as a Catalyst to Decarbonize Global Logistic

16、sW H I T E P A P E RJ A N U A R Y 2 0 2 5In collaboration with McKinsey&Company Transformation of Industries in the Age of AIImpact on industrial ecosystemsCross industryIndustry or function specificImpact on industries,sectors and functionsAdditional reports to be announced.Blueprint to Action:Chin

17、as Path to AI-Powered Industry TransformationW H I T E P A P E RJ A N U A R Y 2 0 2 5In collaboration withAccentureTransformation of Industries in the Age of AIAI Governance AllianceRegional specific Impact on regionsAdvanced manufacturing andsupply chainsFinancial servicesMedia,entertainment andspo

18、rtHealthcareTransportTelecommunicationsConsumer goodsArtificial Intelligence in Media,Entertainment and SportW H I T E P A P E RJ A N U A R Y 2 0 2 5In collaboration withAccentureTransformation of Industries in the Age of AIAI Governance AllianceLeveraging Generative AI for Job Augmentation and Work

19、force ProductivityArtificial Intelligences EnergyParadox:BalancingChallenges andOpportunitiesArtificial Intelligence and Cybersecurity:Balancing Risks andRewardsAI in Action:Beyond Experimentation to Transform IndustryBlueprint to Action:Chinas Path to AI-Powered Industry TransformationArtificial In

20、telligence inFinancial ServicesFrontier Technologies in Industrial Operations:The Riseof Artificial Intelligence AgentsArtificial Intelligence in Media,Entertainment and SportThe Future of AI-Enabled Health:Leading the WayIntelligent Transport,Greener Future:AI as a Catalyst to Decarbonize GlobalLog

21、isticsUpcoming industry report:TelecommunicationsUpcoming industryreport:Consumer goodsFrontier Technologies in Industrial Operations3ForewordAmid a landscape of exponential technological change,society is entering the Intelligent Age1 anera defined by far more than technology alone.The Intelligent

22、Age is characterized by a mass revolution transforming all aspects of society,and were already beginning to witness profound shifts.Alongside technological intelligence,environmental,social and geopolitical intelligence will be fundamental to success in this age.In this new era,industrial operations

23、 are being redefined.To better understand emerging opportunities and explore potential responses,the World Economic Forum in collaboration with Boston Consulting Group(BCG)launched the global initiative Frontier Technologies for Operations:AI and Beyond.Building on the success of our previous AI-Pow

24、ered Industrial Operations initiative from 2022,this new scheme aims to equip manufacturers with the insights and tools necessary to shape the future of industrial operations in the Intelligent Age.Its important to address two pressing questions why focus on frontier technologies,and why now?The ans

25、wer is simple yet profound these innovations,like others in the Intelligent Age,drive boundary-breaking advancements that push the limits of whats currently possible,facilitating collaborative intelligence and amplifying human ingenuity.In doing so,they provide competitive advantages and catalyse su

26、stainable growth.Manufacturers that fail to fully harness the transformative power of frontier technologies in operations and supply chains will surely fall behind.Although the pursuit of frontier technologies is not novel,the stakes are now higher than ever.The challenges associated with identifyin

27、g and evaluating these technologies and integrating them into long-term strategies have grown more complex as the pace of innovation accelerates.Forward-thinking industries,technology leaders and academic institutions are pioneering such advancements.Yet,even with the growing accessibility of tools

28、like generative AI,manufacturers still face a crucial question how will frontier technologies drive real,measurable impact in day-to-day operations?This white paper presents a bold yet actionable vision of one such frontier technology AI agents.It additionally outlines methods by which this technolo

29、gy could be applied to create tangible value in industrial operations.The paper focuses on the transformative potential of two types of AI agents virtual AI agents and embodied AI agents and provides insights and case studies from leading industries while challenging conventional thinking and inspir

30、ing new strategies.Its aim is to highlight innovative perspectives to help manufacturers unlock the full potential of AI agents and spearhead operational transformation.Kiva Allgood Head,Centre for Advanced Manufacturing and Supply Chains;Member,Executive Committee,World Economic ForumDaniel Kpper M

31、anaging Director and Senior Partner,Boston Consulting Group(BCG)Frontier Technologies in Industrial Operations:The Rise of Artificial Intelligence Agents January 2025Frontier Technologies in Industrial Operations4Executive summaryThe manufacturing landscape is becoming increasingly complex,and this

32、trend is projected to accelerate in the coming years.Labour shortages,rising cost pressures and shifting customer demands,geopolitical dynamics and decarbonization goals necessitate significant operational transformation.Current technologies will be insufficient to drive the required levels of flexi

33、bility,sustainability and excellence needed to facilitate this change.To succeed,manufacturers can embrace frontier technologies that push the limits of innovation.However,navigating this rapidly evolving technological landscape is challenging,as many manufacturers need to address immediate operatio

34、nal needs and plan for the future of theiroperations.Industrial operations are likely to evolve towards an artificial intelligence(AI)-centric model,where AI drives self-controlling,near-autonomous systems while empowering humans.While near-autonomous operations may become common in some industries,

35、human involvement will remain crucial.The role of humans will be redefined,with workers transitioning from hands-on operators to orchestrators,stepping in when judgment or creativity is required.This shift will boost operational efficiency,allowing humans to focus on strategic tasks and ethical deci

36、sion-making to drive innovation and growth.In the broad landscape of frontier technologies,AI and more specifically,rapidly evolving AI agents have the potential to propel manufacturers towards this future,and unlock novel opportunities in operations across many industries.This report focuses on two

37、 types of AI agents:virtual AI agents and embodied AI agents.These agents are expected to enhance both digital applications and physical systems,and perform complex tasks with minimal human intervention.Virtual AI agents advancing autonomous software systems:Virtual AI agents enable software applica

38、tions to autonomously achieve defined goals in the digital environment,acting as assistants,advisers or automation agents.These agents support workers and can also independently control and steer processes andmachinery.Embodied AI agents ushering in a new era of robotics:Embodied AI agents equip phy

39、sical systems,such as robots,with the ability to perceive and act within the physical environment,allowing for dynamic and complex movements.These advancements will be crucial for overcoming the current limitations ofroboticautomation.Successfully navigating the transition to near-autonomous,AI-agen

40、t-driven operations requires a comprehensive,value-driven approach to technology adoption.Solutions should be scalable and aligned with long-term business objectives.Establishing strong organizational and technological foundations that support this vision will be crucial for manufacturers looking to

41、 capture the technologys full potential.The insights presented in this paper are focused on manufacturing and founded on the collective expertise of the initiative community,drawing from consultations with senior executives and academic experts.Moving forward,the community will continue to work clos

42、ely with manufacturing stakeholders across industries to deliver a global,comprehensive outlook on the future of industrial operations.This effort will concentrate on recent and future frontier technologies,with an emphasis on responsible transformation approaches.Its essential that manufacturers em

43、brace frontier technologies to secure a thriving,sustainable future in manufacturing.Frontier Technologies in Industrial Operations5IntroductionFrontier technologies have pushed the limits of what is possible in industrial operations over the past decades,significantly boosting productivity,reducing

44、 costs and improving the work environment.Innovations like robotics and the industrial internet of things(IIoT)have been instrumental in modernizing operations and laying the foundation for the next wave of breakthroughs.Today,the technological landscape is evolving at an unprecedented pace.This pro

45、gress is primarily driven by the exponential increases in computing power and breakthroughs in artificial intelligence(AI)society is currently witnessing.As new frontier technologies emerge,manufacturers face the key challenge of discerning which innovations will bring lasting value at scale,and whi

46、ch are merely transient trends.This creates uncertainty around where to focus development efforts and investments.Overcoming these challenges is essential.Harnessing the value of frontier technologies is now vital for manufacturers as they seek to maintain a competitive edge and tackle industry-spec

47、ific obstacles.To retain a leading position in the evolving landscape,companies must not only adopt these innovations but also understand the transformative impact on the future of operations.Success in this journey hinges on answering a few key questions:What will the future of industrial operation

48、s looklike?Where is the real value in this transformation?Which frontier technologies will address keychallenges?What steps need to be taken to realize value atscale?Drawing on insights from experts and executives across operations and technology,this white paper provides a strategic perspective on

49、these questions,with a focus on AI-agent-enabled transformation.It presents a forward-looking vision of AI-driven,near-autonomous industrial operations.It explores the role of AI agents in enabling this vision,specifically virtual AI and embodied AI agents,offering concrete examples and case studies

50、 to demonstrate their value.Additionally,it outlines the strategic imperatives necessary for successfully scaling these technologies.While AI agents hold transformative potential,it is crucial to recognize that they are not yet fully developed.Leading companies are running pilots to test their capab

51、ilities,with their at-scale impact to be realized in the coming years.Although not covered in this white paper,other frontier technologies such as biotechnology and quantum technology are generating significant interest.These technologies hold the potential to revolutionize manufacturing operations,

52、either directly or indirectly,but remain in earlier stages ofdevelopment.AI agents are transforming industrial operations,driving efficiency and unlocking competitive advantages.The two types of AI agentsBOX 1Virtual AI agentsSoftware-based AI agents that operate entirely in the digital environment

53、and enable digital applications to autonomously achieve definedgoalsEmbodied AI agentsAI agents integrated into physical systems such as robots that interact with the physicalenvironment As new frontier technologies emerge,manufacturers face the key challenge of discerning which innovations will bri

54、ng lasting value at scale,and which are merely transienttrends.Frontier Technologies in Industrial Operations6The next leap:reinventingindustrial operations through frontiertechnologies1Manufacturers face a complex operating environment with growing challenges:Cost competitiveness:Rising labour cost

55、s,supply chain disruptions and international competition necessitate improved efficiency andlowered structural costs.Labour shortages:More than 2.1 million manufacturing jobs are projected to remain unfilled in the US alone until 2030,2 driving workforce risks and productivity challenges.Customer de

56、mands:Consumers expectations for greater customization and faster delivery drive the need for more flexible production systems and better demand forecasting.Geopolitical dynamics:Tariffs and fragmented production across multiple geographies hinder economies of scale,leading to greater complexity in

57、supply chains,dispersed know-how and increased risks.Sustainability:To meet decarbonization goals,its crucial to optimize energy and resource use while reducing emissions through robust supply chain management.Addressing these challenges requires a shift in operational excellence,breakthrough innova

58、tion,structural optimization,supply chain diversification and investment in regional manufacturing clusters.Preparing for the challenges ahead requires operational transformation driven by frontier technologies.Frontier Technologies in Industrial Operations71.2 Redefining the role of humans:from ope

59、rators to AI-enabled orchestrators1.1 Entering the next frontier:the path towards self-controlThe industrial sector stands at a pivotal juncture.Frontier technologies,such as AI agents,are capable of performing complex activities.This paves the way for increasingly AI-driven,near-autonomous operatio

60、ns,within which many machines and AI-enabled systems will function with minimal human intervention.Success depends on cultivating a trusted human-machine interaction,where both collaborate seamlessly.Currently,automation is often reserved for simple,repetitive tasks that still require manual oversig

61、ht to ensure continuous operation.In the past,the expansion of automation was hindered by technological hurdles(such as an inability to handle unsorted flexible parts like cables automatically)and financial constraints.However,more advanced technologies and decreasing costs are poised to enable wide

62、r deployment across factories,with autonomous systems taking control of routine operations.These autonomous systems encompassing machines,robots and virtual systems may manage routine tasks ranging from material handling to quality control and production planning.Such systems may optimize and adjust

63、 production parameters on machines in real time to align with business needs,enhancing flexibility.Although the extent of automation will ultimately depend on the return of investment across industries and regions,many factories may converge towards autonomy,driven by the need to remain competitive.

64、The shift towards autonomy may also revolutionize factory design.Future AI-centric factories might prioritize machine-optimized layouts that enhance production efficiency and flexibility.For instance,valuable ground-floor space can be freed up by storing unfinished parts in automated multi-storage s

65、helves,manual processes can be accelerated and performance monitoring can be centralized in virtual control centres rather than dispersed throughout the shop floor.Self-controlling factories and supply chains will deliver significant improvements such as:Efficiency:Predictive analytics will shift op

66、erations from reactive to proactive management,anticipating issues and implementing necessary adjustments immediately.Real-time adjustments will enhance machine uptime,quality control and cost efficiency.Flexibility:Advanced robotics and AI will enable highly personalized manufacturing and swift rec

67、onfigurations,making production lines adaptable to varying product demands.Autonomous systems will self-organize for optimal factory layout and performance,further enhancing flexibility.They will also increase supply chain agility and responsiveness.Sustainability:Autonomous systems will optimize en

68、ergy consumption and minimize waste.Real-time analytics will monitor environmental impacts,ensuring that sustainability goals are met without sacrificingefficiency.Worker empowerment:AI-driven tools and automation will enhance workforce capabilities and facilitate human-machine interactions,enabling

69、 workers to quickly understand production issues and make more well-informed decisions.The transformation to near-autonomous industrial operations requires coordinated changes across both human and technological dimensions.Human involvement will remain essential in industrial operations of the futur

70、e,as workers may transition from hands-on operators to AI-enabled orchestrators who oversee autonomous systems and provide judgment or ingenuity as required.As machines advance in natural language comprehension,human-machine interactions will become more fluid and intuitive,enabling productivity bre

71、akthroughs.For example,one individual supported by assistant systems can supervise multiple functions such as quality,inspection andproduction simultaneously.Maintenance activities that require physical dexterity such as checking for leaks or replacing parts inside a machine may partially remain hum

72、an-led but can be significantly augmented by virtual agents.Although the extent of automation will ultimately depend on the return of investment,many factories may converge towards autonomy,driven by the need to remain competitive.Frontier Technologies in Industrial Operations8Industry example:Shift

73、ing role of technicians and supervisorsIndustry example:Elevating planner roles with AI-supported decision makingBOX 2BOX 3A global wheel manufacturer has experienced a shift in the role of their technicians and supervisors with the introduction of a prescriptive AI solution for process parameter ad

74、justment developed by a Cape Town-based AI solution provider.Instead of managing process details,technicians now focus on identifying root causes and driving continuous improvement by optimizing the plan-do-check-act(PDCA)cycle.Supervisors,in turn,are evolving into AI users,interpreting AI-driven in

75、sights and guiding operators towards more efficient problem-solving.This transition enables both operators and supervisors to concentrate on long-term,systemic improvements rather than routine,reactive tasks.A Fortune 500 technology manufacturer elevated the role of its planners from executors to ar

76、chitects of its supply chain decision-making process.Previously relying solely on humans,the company struggled with delayed decision-making,resulting in large inventories and long lead times.By harnessing an AI agent solution from a US-based decision intelligence company,they automated routine decis

77、ions in inventory management while routing exceptions to human experts with contextual data,analysis and recommendations.The platform optimized stock levels and ensured supply was matched to regional demand.As a result,77%of agent recommendations were automatically executed and 90%were accepted with

78、out change.This evolution will require manufacturers to anticipate a transition in workforce skills and cultural identity,making early engagement of operators in the transformation journey critical for success.In a future with largely self-controlling systems,humans may partner with machines,harness

79、ing collaborative intelligence to focus on higher-value tasks,such as:Strategic decision-making involves using AI-driven recommendations to make business-critical decisions.For instance,in an automotive plant,AI may recommend adjustments to production schedules or shift planning.A human planner may

80、weigh these recommendations against factors such as projected customer demand or current labour availability.Performance supervision involves monitoring and adjusting autonomous systems as needed.For instance,in a semiconductor plant,operators may monitor autonomous systems handling wafer fabricatio

81、n.If performance metrics show yield deviations that systems cannot resolve,humans can step in to address the issue.Continuous improvement involves solving complex problems and optimizing processes.For instance,in a chemical processing plant,engineers may use AI to identify inefficiencies in mixing o

82、r reaction processes.They can then redesign workflows or machine configurations to optimize output and reduce waste.Creativity and innovation involve developing new production processes and rethinking factory layouts.For instance,in a consumer electronics plant,a maintenance worker might introduce c

83、reative ideas to streamline tool changes by mounting additional supports that have been employed in other industries.In a future with largely self-controlling systems,humans may partner with machines,harnessing collaborative intelligence to focus on higher-value tasks.Frontier Technologies in Indust

84、rial Operations9AI agents fuelling the transformation ofoperations2Virtual and embodied AI agents could drive the transition towards near-autonomous operations in both software and robotics.Realizing the transformative vision of AI-centric operations requires a thorough assessment and evaluation of

85、the potential of AI agents.Both virtual and embodied AI agents have the potential to deliver significant value,unlock new opportunities and drive the transition towards near-autonomousoperations.AI will transform from a data-centric front end to an agent-centric user end,relying on domain-specific d

86、ata sources to optimize industrial operations.These domain-based agents will drive new growth of AI across different industries.The interactive agents will further transform the new large knowledge model,fostering the development of AI ecosystems with advanced technologies,tools and talents.Jay Lee,

87、Clark Distinguished Professor;Director,Industrial AI Center,University of Maryland10Frontier Technologies in Industrial OperationsAI agents function in a continuous observe,plan and act cycleExecute by leveraging internal or external tools/systemsActEvaluate possible actions to prioritize them throu

88、gh reasoningPlanCollect and process data from environmentObserve AgentThe basics of AI agentsBOX 4Source:Boston Consulting Group(BCG).AI agents amplify the impact of large language models(LLMs)by giving them access to tools and enhancing their ability to observe,plan and execute actions.3 Traditiona

89、l AI algorithms,such as machine learning,are task-specific and require human input for defining tasks,providing data and interpreting results.In contrast,AI agents,once trained,can operate and achieve specific objectives autonomously,continuously observing their environment,planning actions and harn

90、essing tools to execute complex tasks.AI agents function in a continuous observe,plan and act cycle,which makes them particularly valuable for operations.Each step is enabled by interfaces or modules:4 Observe:Agents collect and process data from the environment,including multimodal data,user input

91、or data from other agents.For example,an agent can perceive deviations in production quality and underlying parameters in real time.Agent-centric interfaces:Agents require protocols,application programming interfaces(APIs)and specifically designed interfaces to input multimodal data or perceive real

92、-time data from multiple sources.Memory module:Agents have short-and long-term memory,which allows them to remember general knowledge,past actions and decision-making.Plan:Agents and their underlying LLMs evaluate possible actions to prioritize them through logical reasoning,in accordance with their

93、 objectives.In the example above,the agent reviews possible actions to improve quality and decides to change production parameters.Profile module:Agents have defined attributes,identities,roles or behavioural patterns.The roles can be predefined,or agents can be flexible and dynamically adapt to new

94、 roles.Reasoning module:Agents have limited reasoning capabilities.The underlying LLM is capable of decomposing theagents prompts and returning an actionable plan.It extracts key insights and makes logical connections by replicating reasoning steps observed in training data.This enables agents to de

95、cide on the required next steps by breaking down complex tasks into small actions to achieve their objectives.Recent studies have shown that current LLMs are not yet capable of formal reasoning.Real-world solutions thus require other types of AI and solvers and cannot solely rely on existing LLMs.5

96、Act:Agents execute actions by harnessing internal or external tools and systems.For example,an agent accesses the machine controller and changes the defined machine parameters.Action module:Agents decide which tools to use,using access mechanisms such as APIs,system integrations or other agents as n

97、eeded.Functioning in this cycle,agents continuously learn from self-reflection or external feedback.Through goal-oriented learning approaches,such as reinforcement learning,agents continuously adapt and refine their strategies over time.This makes them particularly valuable in complex,dynamic enviro

98、nments where conditions and objectives are constantly shifting.Such environments can be found widely across industrial operations.As part of multi-agent systems,in which specialized agents work together by dividing complex problems among themselves,they can automate entire processes end-to-end.Front

99、ier Technologies in Industrial Operations11The four types of virtual AI agentsFIGURE 1Maturity levelSpecialistagentsMeta agentsMeta agentKnowledge agentAssistant(executing manual tasks)Recommendation(proposing scenarios and actionable insights)Automation(autonomously performing activities)Adviser ag

100、entAutomation agentSource:Boston Consulting Group(BCG),World Economic Forum.Knowledge agents support workers as intelligent assistants.They analyse and synthesize vast amounts of data to provide real-time operational insights,flag anomalies and create content such as reports and code.By accessing mu

101、ltiple tools and real-time data sources,such as machine logs and sensor data,they add value to functions that require quick insights for example,in maintenance,quality and logistics.They can also support engineering with machine code generation.Adviser agents go further by generating real-time scena

102、rios to address issues,and recommending actionable insights.They continually refine their recommendations based on real-time feedback,enabling them to learn autonomously and adjust actions such as machine parameter setting,workforce management,production planning and factory layout optimization.They

103、 also suggest the best possible scenario based on their optimization objective and received feedback,empowering users to align decisions with business priorities.Automation agents act independently,executing optimal actions without human input.They adapt to new situations through real-time feedback

104、without explicit retraining,allowing them to autonomously optimize machine performance,adjust production parameters,recode instructions or modify production plans.They surpass existing RPA(robotic process automation)by automating not only individual tasks but also entire human activities that requir

105、e understanding,planning and execution.Meta agents orchestrate specialist agents in the context of multi-agent systems to achieve broader objectives,enabling area-or even factory-wide steering.The long-term vision for meta agents is to consolidate knowledge and automate end-to-end supply chains by i

106、ntegrating diverse specialized agents.Within a factory,these agents could cover an entire production process or group of machines.While specialist agents are already being piloted across industries,meta agents require enterprise-wide AI and further development before real-lifeimplementation.Virtual

107、AI has a significant impact across all manufacturing and supply chain functions,from logistics to production,as well as support functions such as maintenance,quality and engineering.The two use cases described below production process parameter setting and real-time production planning illustrate th

108、e agents capabilities.2.1 Virtual AI paving the way for autonomous systemsVirtual AI agents can manage a wide range of software-based tasks,from routine operations and research to advanced analytics and task automation.In industrial operations,they can enhance responsiveness,improve execution qualit

109、y,boost productivity and reduce operational mistakes.Unlike traditional machine learning programmes,they can make context-sensitive decisions in real time and adapt through feedback loops.These agents have applications across all operation functions,including production,maintenance,quality,engineeri

110、ng,logistics and planning.The maturity of virtual AI agents can be categorized into three levels:assistant,recommendation and automation.The distinct objectives at each maturity level are pursued by specialist agents:Virtual AI agents have applications across all operation functions,including produc

111、tion,maintenance,quality,engineering,logistics and planning.Frontier Technologies in Industrial Operations12Application of virtual AI agents in production process parameters setting and real-time production planningTABLE 1Use case description1 Production process parameters settingAchieving optimal s

112、etting of machine parameters(such as temperature and pressure)is a key goal for manufacturers across all industries.The complexity of optimizing various process parameters under specific external influencing factors forces manufacturers to rely heavily on operator experience.AI agents are transformi

113、ng this process to improve overall equipmenteffectiveness.2 Real-time production planningReal-time production planning is critical for manufacturers to meet demand,reduce lead times and optimize resource use.However,the complexity of balancing capacity,inventory,labour and external factors often req

114、uires manual adjustments by experienced planners.AI agents are transforming this process by streamlining decision-making and improving flexibility and responsiveness to changing conditions ultimately enhancing overall production efficiency.Knowledge agentParameter knowledge agents harness machine an

115、d process parameters and production output data(such as quality validation,material properties,and maintenance and quality reports)to identify optimal equipment settings for enhanced machine performance.The agent can be activated by workers via voice input or can raise alerts when predicting deviati

116、ons.It continuously refines its analysis based on worker feedback.It also assists decision-making by estimating the potential value and cost impact of potential adjustments.Planning knowledge agents serve as foundational tools that gather and synthesize data from multiple sources,such as historical

117、production performance,demand forecasts,inventory levels and resource availability.By harnessing this data,an agent provides planners with actionable insights.It can analyse past trends to identify potential bottlenecks,recommend best practices and highlight opportunities to streamline production pr

118、ocesses.Planners can query the agent to retrieve detailed analysis or predictive insights,making it a valuable decision-support tool.Over time,it refines its knowledge,continuously improving the accuracy andrelevance of its insights.Adviser agentParameter adviser agents continuously monitor machine

119、performance,detect real-time deviations and recommend setpoint improvements to achieve desired production goals.Operators can validate the recommended settings and corrective actions.The agent refines its recommendations over time by integrating real-time worker feedback and assessing performance ou

120、tcomes.Planning adviser agents build on the capabilities of the knowledge agent by continuously monitoring real-time production and forecast data.They detect potential deviations,such as delays,resource shortages or equipment downtime,and recommend proactive adjustments to the production plan.These

121、recommendations can include rescheduling,resource reallocation or inventory adjustments to ensure operational targets are met.The agent learns from the planners validation and feedback,allowing it to improve its predictions and better anticipate future challenges.This adaptability augments its abili

122、ty to optimize production planning.Automation agentParameter setpoint automation agents continuously track machine performance,identify anomalies in real time and autonomously adjusts setpoints or take corrective actions.They adapt and self-correct parameters based on current production priorities w

123、ithout requiring human intervention.The agent can also work with digital twins,incorporating necessary external inputs such as customer demands into the digital twin to support business decisions.Although pilots in series production are still pending,a research study has demonstrated how LLM agents

124、can control and steer operations remotely,either with human oversight or through AI agents in a digital twin.6Planning automation agents take the next step by autonomously managing production schedules in real time.They respond dynamically to production events like machine breakdowns,labour shortage

125、s or fluctuations in demand.The agent continuously updates the production plan,reallocates resources and reschedules tasks to ensure the overall production process remains efficient.Unlike the adviser agent,the automation agent acts independently to adjust,only requiring human oversight for major ex

126、ceptions or strategic decisions.Meta agentMeta parameter agents oversee and orchestrate multiple dedicated machine parameter agents,synchronizing setpoints across various machines.Thisagent autonomously coordinates actions to ensure an optimized production flow,preventing bottlenecks and dynamically

127、 adjusting the system to optimize overall performance.Meta supply chain agent oversees multiple planning agents that focus on specific areas such as labour,inventory,or capacity.It coordinates these agents to ensure that the entire production process remains balanced and optimized.By dynamically adj

128、usting priorities and resources across various departments,the meta agent ensures that the production plan aligns with broader organizational goals and avoids conflicts between localized planning decisions.Frontier Technologies in Industrial Operations13Snapshot of sample case studies of virtual AI

129、agentsTABLE 2Use caseChallengeSolutionBenefitsAutonomous control agent for steel manufacturerSteel manufacturer KG Steel faced two main challenges:1)high liquified natural gas(LNG)energy costs to operate furnaces and,2)discrepancies in the product quality arising from a skill gap caused by an ageing

130、 workforce.To meet the desired quality of coils,operators must adjust the heating settings of furnaces while accounting for varying production environments.Inefficient heating process control causes excessive use of LNG.A South Korea-based AI software provider developed a deep learning model predict

131、ive control optimization model that executes“what-if”scenarios based on possible control patterns,and assesses them in a digital twin to select the optimal controls.Initially,the agent acted as adviser,providing operators with recommendations on optimal control settings.Partial automation of furnace

132、 operations was later achieved via system integration and direct feeding of agent output into the furnace control system.The agent has decreased LNG consumption by approximately 2%,while reducing differences in the product quality.Planning AI agent for global brewerA global brewer aimed to improve i

133、ts planning process and forecasts.The current approach,known as post-game analysis,entailed continuously assessing past forecast inaccuracies and their root causes.This was tedious and required expert knowledge that dissipates over time.Post-game analysis is increasingly conducted by LLM composite a

134、gents that integrate a sequence of atomic agents to perform complicated exercises.They are trained on post-game“recipes”built by planning experts,learn continuously from feedback to improve results over time and can be used in cross-functional planning processes.Beyond productivity,the agents enhanc

135、e knowledge and expertise in an organization.By using the agents of a US-based planning solution provider,the brewer achieved 70%touchless demand and supply planning.They additionally issued a resolution for feasibility checks.This percentage is expected to increase further as complex LLM/agent capa

136、bilities are added and further enhanced.Key success factors for adoption and scaling are the quality of the data,the ability to capture the decision-making logic of planners,plan explainability and trustworthiness of outcomes.Based on these criteria,the mid-term ambition is to achieve 90%automation

137、scope and global scaling across markets.Autonomous quality control AI agentAs a globally recognized digital lighthouse,7 Siemens Electronics Work Amberg(EWA)has set the ambitious goal of achieving a first pass yield(FPY)exceeding 95%with a defect per million connections(DPMC)below 10 per production

138、batch.This is a significant challenge,given that a circuit board can have up to 3,800 quality features to monitor.So far,the FPY target has been unachievable as employees have not been able to consistently make the right decisions given the time constraint and stress.To achieve this objective,EWA de

139、veloped a patented autonomous quality control AI agent in collaboration with RIF Institute for Research and Transfer.This advanced agent,harnessing self-organizing maps,assists employees in correctly setting up the solder paste printer for the first production run.This reduces process times in a com

140、plex,multi-parameter task that typically requires significant experience.As process parameters are continuously improved,the agent adapts by accounting for parameter changes,resulting process behaviour and prior adjustments.This enables it to continuously learn and optimize the unknown behaviours of

141、 the solder printing process.After the testing phase,the agent will be able to autonomously adjust the parameter settings.Multiple studies have consistently demonstrated high-quality product output while simultaneously reducing the solder paste printers process time by up to 50%compared to the cycle

142、 time.The next phase of development involves transforming the agent into an edge application for rapid scalability.A key requirement for this advancement is integrating the quality inspection gates into the comprehensive digital twin of the process.Frontier Technologies in Industrial Operations142.2

143、 Embodied AI igniting a new era in roboticsAI is not only transforming software but also automating physical workflows.Embodied AI integrates AI into physical systems such as robots,allowing them to perceive and interact with their environment through dynamic and complex movements.The agents see the

144、 world via sensors(for example,cameras,radar,lidar and microphones)and execute actions through actuators such as advanced grippers.Applied to industrial operations,these agents enhance the capabilities of existing robotic systems,enabling more sophisticated automation.By doing so,they expand the aut

145、omation scope,overcoming traditional challenges such as those associated with handling unstructured environments or manipulating unstable objects.Three types of robotic systemsFIGURE 2Robotic software improvementRobotic hardware improvementRobot capabilities enabled by embodied AIRule-based robotics

146、CodingPerformance and reliabilityTraining-based roboticsTrainingTask versatility and flexibilitySituation adaptabilityManipulation dexterityContext-based roboticsZero-shot learningGeneral understanding and task executionHuman-like dexterity and low-level controlUniversal robotic embodimentSource:Bos

147、ton Consulting Group.Notably,three types of robotic systems have emerged:rule-based,training-based and context-based(Figure 2).This evolution has been driven by improvements in both robot hardware and software.The hardware is becoming more capable,reliable and flexible.At the same time,the software

148、is advancing,with improvements in foundation models and technologies(such as reinforcement learning).8 A five-fingered robotic hand with 24 of freedom can perform complex tasks with an unprecedented level of dexterity.9 This is made possible by the various data sources that can be harnessed to train

149、 AI-enabled robots:Real robot data:This data is collected from the robot motion controllers and can also be generated by human-guided robot teleoperation.Although real robot data is the most accurate,it is limited because it can only be gathered from deployed robot fleets.Synthetic robot data:This d

150、ata is created in simulated physics-based environments and is available in infinite supply.While any scenario can be simulated,a simulation-to-reality gap is expected to remain due to the diversity of the real world.This means real robot data will still be necessary for validation.For example,Foxcon

151、n trains robots in its virtual factory,using digital twins to generate synthetic data for model training and to teach robotic arms how to see,grasp and move objects.10 Internet-scale human data:Online data,including human videos,is highly diverse,and equips AI with a foundation for understanding the

152、 world.It provides valuable information on how humans interact with objects and how objects behave.Imitation learning allows the latest models to learn these skills by mimicking human actions.The discussion of the three robot types enabled by embodied AI centres on these software advancements,which

153、harness advanced datasets:Rule-based robotics:Beginning in the 1960s,industrial robots operated under rule-based systems,following“if then”instructions that were manually coded by experienced robotic engineers.Complex Embodied AI integrates AI into physical systems such as robots,allowing them to pe

154、rceive and interact with their environment through dynamic and complex movements.Frontier Technologies in Industrial Operations15automation solutions required individual programming of each robot.These robots were limited to simple,repetitive tasks,allowing for minimal flexibility.Training-based rob

155、otics:Embodied AI is a major technological breakthrough in robotics.The convergence of robotics,machine learning types such as reinforcement learning(RL),and advanced vision systems has transformed automation.By giving robots an understanding of the world,embodied AI enables new applications like bi

156、n picking.Unlike rule-based systems that depend on manual coding,AI-enabled,training-based robotics can now learn skills via RL in a trial-and-error approach in physical or simulated environments.Context-based robotics:Context-based,autonomous robotics are built on robotics foundation models(RFMs)an

157、d have a general understanding of the world.11 Because they require neither coding nor training by manufacturers,they can lead to a paradigm shift in robotics that is,zero-shot learning.This considerably reduces the effort required to programme or teach these robots,opening the way to new,highly com

158、plex applications such as handling cables and addressing unforeseen events.RFMs are still in the development phase and are expected to break through in the coming years.Benefits and examples of the different robot types enabled by embodied AITABLE 3Robot typeBenefitsExampleRule-based robotics Perfor

159、mance and reliability:Robots execute repetitive tasks with defined precision and speed,based on their coded robot programme.The vast majority of todays global robot fleet is rule-based,including industrial robots in assembly lines and automated guided vehicles(AGVs)in logistics.For repetitive tasks

160、that do not entail deviation or special requirements,such as automated assembly lines of medical products,rule-based robots will likely stay the norm in the future.Training-based robotics Task versatility and flexibility:Robots understand the manufacturing world they have been trained in,giving them

161、 the necessary flexibility to adapt to different environments and the dexterity to handle known objects and perform versatile tasks.For example,a kitting robot can now handle a large variety of distinct parts,applying its learned skill library.Situation adaptability:Robots can autonomously understan

162、d and solve unforeseen events in their trained domain by executing the required actions.For instance,if a screw gets stuck,robots can resolve the issue independently,significantly enhancing system reliability.Beyond articulated robots,embodied AI models enhance the capabilities of autonomous mobile

163、robots(AMRs)for material transport drones and other robot types.Manipulation dexterity:Robots can learn to move objects with dexterity,enabling them to conduct advanced movements such as contact-rich assembly of multiple gears,for example.Kitting robots in a warehouse can handle a diverse range of p

164、arts of varying dimensions and characteristics.Previously,kitting operations required manual intervention or multiple robots to manage different part sizes,lowering the return on investment.For example,Otto Group,an online retailer founded in Europe,has deployed AI-controlled robots in its fulfilmen

165、t centre to handle the order-picking process.Thanks to its AI capabilities,the robot can process a wide variety of shapes,colours and quantities,which previously required human hand-eye coordination for items such as textiles.12Frontier Technologies in Industrial Operations16The effect of virtual an

166、d embodied agents on industrial operations will be a tectonic shift,recasting how AI agent systems are built and function.However,as the futurist Paul Saffo noted,Never mistake a clear view for a short distance.We are still in the infancy stage of understanding these agents potential.Generative AI i

167、s not intelligence in a box it doesnt truly reason and cant solve complex optimization problems alone.To safely deploy robots in dynamic,human-inhabited environments,combining large language models,vision language action models and other AI modalities along with engineering is essential,ensuring pro

168、per guardrails are in place for reliable,effectivesolutions.Anthony Jules,Co-Founder and Chief Executive Officer,Robust.AIManufacturing companies are beginning to integrate virtual and embodied AI agents into their operations,as showcased by the pilots discussed above.Although AI agents are still in

169、 the early stages of development and require further refinement for large-scale deployment,itis essentialfor industry leaders to consider the foundational elements necessary for supporting their successful implementation and integration intoexisting systems.Benefits and examples of the different rob

170、ot types enabled by embodied AI(continued)TABLE 3Robot typeBenefitsExampleContext-based robotics General understanding and task execution:Robots harness their general understanding of the world to autonomously generate any action and perform any task according to the situation.They can reason,plan a

171、nd act based on the instructions received and the environment.Additionally,they can receive natural language instructions from technicians for example,when asking for validation on how to grasp a particular object.This significantly improves finetuning in the factory and reduces the required skillse

172、t.Human-like dexterity and low-level control:The models enable intuitive movements(for example,opening an object)and fast low-level control,adapting to any situation.They are expected to be able to handle flexible parts,such as cables,and cope with complex physics,such as moving liquid in aclosedcon

173、tainer.Universal robotic embodiment:General RFM can embody any robot form,including humanoids,articulated robots or mobile robots,allowing for universal use of the models.The development of RFMs will be a breakthrough in robotic capabilities,especially for humanoid robotics.This is due to the large

174、amount of human data that already exists to facilitate training of the human form factor.Additionally,because the world is built for humans,these robots will have a huge array of potential applications beyond operations.Humanoid robots are a prominent example of context-based robotics,which has emer

175、ged in recent years.BMW is piloting the use of humanoid robots for assembly preparation in its Spartanburg plant.13 The human form factor brings multiple benefits that can also be applied to operations,such as in the reuse of existing workstations.The future of humanoids in operations remains uncert

176、ain,however,as their human shape drives significant complexity that is not needed on a factory shop floor,such as bi-pedal locomotion instead of wheels.Even so,the underlying RFM will bring a breakthrough in operations,as it can embody any robotic systems,regardless of form.Frontier Technologies in

177、Industrial Operations17Strategic imperatives for industrial operations transformation3Transformation success hinges on a value-driven,end-to-end approach,grounded in strong organizational and technological foundations.The crucial considerations and steps of the AI journey described in Harnessing the

178、 AI Revolution in Industrial Operations:A Guidebook,published in October 2023,provide the starting point for manufacturers preparing to implement AI technologies at scale.However,the scope of considerations for frontier technologies,including AI agents,extends beyond those highlighted in the guidebo

179、ok.To enhance industrial operations and meet new standards,manufacturers need to remain at the forefront of frontier innovations,prioritize value-driven strategies and develop robust organizational and technological foundations.3.1 Paving the way for successful use of AI agents in industrial operati

180、onsThe transformation journey begins with the development of a clear vision for the future of operations thats aligned with the organizations long-term objectives.The journeys success depends on how well the reason behind the transformation is articulated.Although AI agents and frontier technologies

181、 are provoking great excitement,it is essential to avoid being captivated by technological advancements alone.Technology,by itself,will not generate value and is only one of several success factors for sustainable transformation.For example,embodied AI-enabled automation solutions should be pursued

182、only if they bring a significant return on investment or can contribute to larger applications.Successful manufacturers have adopted a value-driven,end-to-end perspective.Furthermore,while proving value quickly is important to build trust within the organization,ensuring the scalability ofthe soluti

183、ons from the outset is crucial.Frontier Technologies in Industrial Operations183.3 Building the foundations:organizational and technologicalStrong organizational and technological foundations are essential for a successful operational transformation and the use of AI agents at scale.Organizational f

184、oundations Governance:Tailoring the organizational structure and operational processes to a companys specific needs is crucial to support the transformation.To fully capture the value of AI agents,companies must review and adapt current processes and work procedures.Skills and capabilities:To unlock

185、 the full potential of AI agents,companies will need abroad set of skills and capabilities.This may include specialized expertise(such as prompt engineering)that the organization currently lacks and may need to build from the ground up.Additionally,to optimize AI agent use,companies must recognize a

186、nd understand the technologys limitations.Supporting employees through upskilling and reskilling opportunities is also key,allowing them to grow along with the company.As these shifts elicit cultural changes,paying attention to employees emotional skills may further facilitate a smooth transition.Ch

187、ange management:To succeed,change management must be led from the top of the organization,engaging all workers in the transformation.Building trust in the selected technologies through transparency and open communication early in the process is crucial to cultivating a culture of change.Developments

188、 in work practices and cultures influence how new and old roles,as well as new and existing talent pools,collaborate.Maintaining a continuous improvement mindset throughout the process ensures adoption and value realization.Ecosystem partnering:Collaboration is an essential enabler for success,due t

189、o each technologys inherent complexity and the depth of expertise required.The challenge is magnified when dealing with technologies that lack commonalities.For instance,embodied AI and virtual AI require different types of expertise and partnerships.Legal compliance:As the regulatory landscape matu

190、res,manufacturers must ensure that AI and other technologies are used responsibly.They must carefully assess associated risks before implementation.For example,AI agents could grant access to data that should remain restricted,so companies must evaluate the potential risks of making certain informat

191、ion widely available.Frontier technologies,especially AI,help companies to enable tremendous value.To stay at the forefront,organizations must embark now on a journey from digital toadaptive towards autonomous operations.Success in this transition requires a well-defined strategy for a clean,digital

192、 core and data foundation,and a transformative organizational mindset,converging the traditional silos of operations.This will not only lead to a competitive advantage but also prepare companies forthe major challenges of our time.Dominik Metzger,Senior Vice-President and Global Head,SoftwareEnginee

193、ring Digital Supply Chain,SAP3.2 Staying at the forefront of AI agent innovationsThe frontier technologies poised to disrupt operations are rapidly evolving.The field of AI agents is still developing,with new capabilities expected in the coming years.Maintaining leadership in the deployment of these

194、 innovations requires a sustained,systematic effort.Identifying and evaluating AI-agent innovations should be an organization-wide endeavour,as their impact extends beyond operations to areas such as engineering,procurement and IT,as well as the overarching business strategy.Companies should apply a

195、 systematic approach to reviewing such new technologies and assessing the maturity of their current operations,to facilitate the preparation,implementation and scaling of processes.The field of AI agents is still developing,with new capabilities expected in the coming years.Maintaining leadership in

196、 thedeployment ofthese innovations requires a sustained,systematic effort.Frontier Technologies in Industrial Operations19Technological foundationsManufacturers should stay ahead in the development of an enabling technology foundation.Achieving convergence of IT(information technology)and OT(operati

197、onal technology)will be crucial to unlocking the full potential of AI.While these enabling technologies form the foundation of modern automation and digital transformation,they are not standalone value drivers.Key technological foundations include:Data sourcing and processing:AI agent applications r

198、equire readily available and accessible data in the right format,appropriate data-processing infrastructure anddata governance.Applications and user interfaces:User-friendly AI agent interfaces are critical for promoting adoption by operators.Interfaces should be intuitive and developed in collabora

199、tion with operators to maximize usability andengagement.High-performance computing:AI applications and advancements in simulation require significant computing power for timely data processing.Large-scale model training is typically conducted on cloud platforms that are commoditized by hyper-scalers

200、.The emergence of on-premises edge computing has reduced traffic to the cloud by enabling computation directly on the factory floor,thereby enhancing energy efficiency and reducing latency in AIapplications.Connectivity:Access to real-time data is crucial for highly automated environments including

201、digital and robotic applications.This is achievable through advanced wired or wireless networks such as 5G networks.Cybersecurity:With end-to-end digitization,AI-enabled cybersecurity has become increasingly important for securing data and IT/OT systems,and combat agent-based threats.Companies need

202、a comprehensive cybersecurity strategy to protect sensitive data and intellectual property in operations.Manufacturers should stay ahead in the development of an enabling technology foundation.Achieving convergence ofIT and OT will becrucial to unlocking the full potential of AIFrontier Technologies

203、 in Industrial Operations20ConclusionThe impact of AI agents is expected to be significant in industrial operations.Recognizing this potential,leading manufacturers are conducting state-of-the-art pilots.The investments and early outcomes of these pioneering efforts highlight the tremendous transfor

204、mative potential of AI agents.They are not only enhancing efficiency but also fundamentally reshaping the competitive landscape of global industries.In the near future,they are likely to become an essential foundation of most factories worldwide,offering manufacturers a pathway to AI-driven,near-aut

205、onomous operations.While this white paper focuses on AI agents of two types virtual AI and embodied AI additional frontier technologies are expected to reshape operations.Additive manufacturing,quantum technology,biotechnology and enabling technologies such as 5G and edge computing are already on th

206、e horizon.To unlock the potential,manufacturers must form pragmatic insights into the comprehensive value impact of these innovations.Along with their promise,AI agents bring challenges related to security,compliance,social responsibilities and infrastructure requirements.A collaborative approach am

207、ong stakeholders including business,academia and policy-makers to exchange best practices and share insights is crucial for responsible transformation.The aim of this document is to encourage industry leaders to take a holistic approach to assessing the value impact of AI agents and other frontier t

208、echnologies in their operations,and to prioritize responsible adoption.Looking ahead,the World Economic Forum will continue to shed light on the latest frontier technologies and innovations in industrial operations,while also providing a unique platform for collaborations and experimentation among i

209、ndustry leaders,technology experts and academics.These collaborations will explore the maturity and value potential of frontier technologies at scale across many industries and end-to-end value chain functions,to unlock value for companies,society and the environment.Frontier Technologies in Industr

210、ial Operations21ContributorsAcknowledgementsThe World Economic Forum thanks the following individuals for their contributions and participation in working groups,interviews and community discussions:Basma AlBuhairanManaging Director,Centre for the Fourth Industrial Revolution Saudi ArabiaMeshal Alma

211、shariDirector,Digital and IT Strategy and Investment,Saudi AramcoIbrahim AlShunaifiIIoT and SMEs SandboxLead,Centre for the Fourth Industrial Revolution,Saudi ArabiaPierluigi AstorinoChief Operating Officer,Ariston GroupGunter BeitingerSenior Vice-President,Manufacturing;Head,Factory Digitalization,

212、SiemensBerk BirandChief Executive Officer,Fero LabsKlaus BlmVice President,Group IT Products&Architecture,Volkswagen GroupAref BoualwanChief Initiatives&Startups Officer,Consolidated Contractors Company(CCC)Michelangelo CanzoneriGlobal Head,Group Smart Manufacturing,Merck GroupWoai Sheng ChowVice Pr

213、esident/General Manager,Global Instrument Manufacturing,AgilentAleksander CiszekChief Executive Officer,3YOURMINDJason Clark Global Vice President of Manufacturing,ExigerHaldun DingecDirector,Production Technologies,ARCELIKLauren DunfordChief Executive Officer,GuidewheelEfe ErdemChief Technical and

214、Commercial Officer,MESS Technology Platform,Turkish Employers Association of Metal Industries(MESS)Sercan EsenCo-Founder and Chief Executive Officer,IntenseyeLead authorsYannick BastubbePrincipal,Boston Consulting Group;Project Fellow,Future Frontier Technologies:AI&Beyond,World Economic ForumDevend

215、ra JainLead,Frontier Technologies for Operations,WorldEconomic ForumFederico TortiLead,Advanced Manufacturing and Supply Chain,World Economic ForumProject teamMaria BassoDigital Technologies Portfolio Manager,Centre for the Fourth Industrial Revolution,Digital TechnologiesMemia FendriContent Curatio

216、n and Operational Excellence Lead,Advanced Manufacturing and Supply ChainsBenedikt GiegerProject Fellow,AI Transformation of Industries,SAPJill HoangInitiatives Lead,AI and Digital Technologies,Centre for the Fourth Industrial Revolution,DigitalTechnologiesBenjamin SchnfuInitiatives Specialist,Advan

217、ced Manufacturing andSupply ChainsFrontier Technologies in Industrial Operations22Saman FaridChief Executive Officer,Formic TechnologiesPaul FarrellExecutive Vice-President and Chief Strategy Officer,BorgWarnerVijay FernandesDeputy General Manager Manufacturing,The Kuwaiti Danish Dairy Company(KCSC)

218、Ric FulopChief Executive Officer and Co-Founder,Desktop MetalDavid GarfieldGlobal Head,Industries,AlixPartnersKai GoebelDirector,Intelligent Systems Lab,Stanford Research InstitutePuay Guan GohAssociate Professor,Academic Director MSc in Industry 4.0,National University of Singapore Business SchoolJ

219、uergen GrotepassChief Strategy Officer,Manufacturing Europe,Huawei Prasanna GururajanIT Senior Director and Manufacturing Product Line Leader,Johnson&JohnsonStijn-Pieter van HoutenSenior Vice President,Consumer Products;Global Knowledge Innovation Lead,o9 SolutionsLuke HuCo-Founder,ElectroderChristi

220、an HuberHead of Digital Hub,Oerlikon ManagementCynthia HutchisonChief Executive Officer,US Centre for Advanced ManufacturingDouglas Johnson-PoensgenFounder and Chief Executive Officer,CirculorAnthony JulesCo-Founder and Chief Executive Officer,Robust.AIJackie JungVice-President,Global Operations Str

221、ategy and Corporate Sustainability,Western Digital CorporationGys KappersChief Executive Officer,DataProphetHagsoo KimManager,Strategic Planning and Business Development,MakinaRocksRam KuppuswamyChief Procurement Officer,Hero GroupFred LaluyauxPresident and Chief Executive Officer,AeraTechnologyAnan

222、d Laxshmivarahan RGroup Chief Digital and Information Officer,JubilantBhartia GroupJay LeeClark Distinguished Professor;Director,Industrial AI Center,University of MarylandNatan LinderCo-Founder and Chief Executive Officer,TulipInterfacesDarko MatovskiFounder and Chief Executive Officer,causaLensTor

223、bjrn NetlandProfessor in Production and Operations Management,Eidgenssische Technische Hochschule Zric(ETH Zurich)Ni JunChief Manufacturing Officer,Contemporary Amperex Technology(CATL)Meirav OrenExecutive Chairwoman and Co-Founder,VersatilePriyadarshi PandaChief Executive Officer,International Batt

224、ery CompanyCyril PerducatSenior Vice-President and Chief Technology Officer,Rockwell AutomationJoris PoortFounder and Chief Executive Officer,RescalePrevlen RambaleeDirector,HCL TechnologiesNashaat Salman Director,Global Manufacturing Strategy&Quality,HitachiAnubhav SinghVice President,Data and Anal

225、ytics,Global Supply Chain,Schneider Electric Jon SobelChief Executive Officer and Co-Founder,Sight MachineSridhar SudarsanChief Technology Officer,AvathonYousufunnisa Syed Director,New Product Innovation(iPhone Operations),AppleFrontier Technologies in Industrial Operations23Amogh UmbarkarVice-Presi

226、dent,SAP Product Engineering,SAPAndre YoonChief Executive Officer,MakinaRocksCong YuVice-President,Engineering,Celonis AI,CelonisSpecial thanks to o9 Solutions,Aera Technology,DataProphet,MakinaRocks and Siemens for their efforts in contributing use cases to this report.ProductionLouis ChaplinEditor

227、,Studio MikoLaurence DenmarkCreative Director,Studio MikoCharlotte IvanyDesigner,Studio MikoFrontier Technologies in Industrial Operations241.Schwab,K.(2024).The Intelligent Age:A time for cooperation.World Economic Forum.https:/www.weforum.org/stories/2024/09/the-intelligent-age-a-time-of-cooperati

228、on/.2.The Manufacturing Institute.(n.d.).Creating Pathways for Tomorrows Workforce Today:Beyond Reskilling in Manufacturing.https:/themanufacturinginstitute.org/research/creating-pathways-for-tomorrows-workforce-today-beyond-reskilling-in-manufacturing/#access-the-report.3.Burtsev,M.et al.(2023).GPT

229、 Was Just the Beginning.Here Come Autonomous Agents.Boston Consulting Group.https:/ al.(2024).A survey on large language model based autonomous agents.Frontiers of Computer Science,vol.18.https:/doi.org/10.1007/s11704-024-40231-1.5.Mirzadeh,I.et al.(2024).GSM-Symbolic:Understanding the limitations o

230、f mathematical reasoning in large language models.https:/doi.org/10.48550/arXiv.2410.05229.6.Jazdi,N.et al.(2023)Towards autonomous system:Flexible modular production system enhanced with large language model agents.Arxiv.https:/arxiv.org/abs/2304.14721.7.World Economic Forum.(13 January 2023).Again

231、st Economic Headwinds,18 Manufacturing Lighthouses Show How to Boost Productivity and Sustainability by Scaling Advanced Technologies across Networks Press release.https:/www.weforum.org/press/2023/01/against-economic-headwinds-18-manufacturing-lighthouses-show-how-to-boost-productivity-and-sustaina

232、bility-by-scaling-advanced-technologies-across-networks/.8.Tang,C.et al.(2024).Deep reinforcement learning for robotics:A survey of real-world successes.Arxiv.https:/www.arxiv.org/abs/2408.035399.Shadow Robot.(n.d.).Dexterous Hand Series.https:/ Trains Robots,Streamlines Assembly With NVIDIA AI and

233、Omniverse.NVIDIA blog.https:/ al.(2023).Towards general-purpose robots via foundation models:A survey and meta-analysis.arXiv.https:/arxiv.org/abs/2312.08782.12.Otto Group.(2024).Strong together:Welcome,robot colleague!.https:/ group.(2024).Successful test of humanoid robots at BMW Group Plant Spart

234、anburg.https:/www.bmwgroup- Technologies in Industrial Operations25World Economic Forum9193 route de la CapiteCH-1223 Cologny/GenevaSwitzerland Tel.:+41(0)22 869 1212Fax:+41(0)22 786 2744contactweforum.orgwww.weforum.orgThe World Economic Forum,committed to improving the state of the world,is the International Organization for Public-Private Cooperation.The Forum engages the foremost political,business and other leaders of society to shape global,regional and industry agendas.