1、 Perspective Accelerating the adoption of telco AI to deliver autonomous networks April 2023 Adaora Okeleke Accelerating the adoption of telco AI to deliver autonomous networks|i Analysys Mason Limited 2023 Contents Contents 1.Executive summary 1 2.Recommendations 3 3.CSPs are adopting AI on their p

2、ath to deploying autonomous networks,but data access is the top challenge 4 Growing revenue,and improving service quality and customer experience are among the strategic objectives driving CSPs investment in AI 4 CSPs should focus on telco AI use cases that create value for their networks and their

3、customers 6 CSPs main obstacle to AI adoption is their lack of access to high-quality data 7 4.CSPs need a clear blueprint for deploying AI uses case and the right vendor partnerships to move to autonomous networks 9 CSPs should adopt an AI use case implementation blueprint,which should be driven by

4、 the ease of access to data and the time that it takes to derive value 9 CSPs should build partnerships with vendors that have combined expertise in telecoms,AI and software 11 5.Conclusion 13 6.About the author 14 List of figures Figure 1.1:CSPs current levels of automation.2 Figure 2.1:Two-stage b

5、lueprint for accelerating telco AI use case deployments.3 Figure 3.1:CSPs strategic business objectives for AI.5 Figure 3.2:CSPs telco AI use case categories.6 Figure 3.3:Top-ranked challenges that CSPs face when developing AI use cases.7 Figure 3.4:Factors that influence the timeline for delivering

6、 AI use cases.8 Figure 4.1:Factors that CSPs should consider when selecting which AI use cases to deploy.10 Figure 4.2:CSPs top vendor selection criteria for AI solutions.12 This perspective was commissioned by Nokia.Usage is subject to the terms and conditions in our copyright notice.Analysys Mason

7、 does not endorse any of the vendors products or services.Accelerating the adoption of telco AI to deliver autonomous networks|1 Analysys Mason Limited 2023 1:Executive summary 1.Executive summary Communication service providers(CSPs)want to accelerate the migration to autonomous networks in order t

8、o address mounting pressures to manage costs,grow revenue and improve customer experience.CSPs need to automate network and service operations as much as possible to grow revenue margins and maximise the return on their investments.Analysys Masons research indicates that USD125 billion of CSP capex

9、investment was spent on building 5G networks in 2022 and we expect a further USD132 billion to be spent in 2023.1 CSPs that can automate their processes can also address emerging sustainable demands such as energy consumption and other associated metrics.CSPs are therefore investing in automation to

10、 drive the improved efficiency of their network infrastructure and operations.Telco AI2 has also been a key facilitator of automation in the telecoms sector.For example,several CSPs are using AI models to analyse large volumes of data.The insights that these CSPs obtain via these models are used to

11、trigger automated functions that support activities such as network planning and optimisation,customer care/service and marketing.Unfortunately,CSPs are unable to access high-quality data sets(which will enable them to make more-accurate decisions)because they are using legacy systems with proprieta

12、ry interfaces.This will restrict how quickly they can integrate AI into their networks.CSPs need to examine their approach to implementing AI so that they can generate value as quickly as possible.They also need to identify ways to improve data access to enable more AI use cases,which will help to m

13、ake operations more efficient and will help to capture new revenue opportunities.Analysys Mason conducted a survey of 84 senior CSP executives worldwide between September and November 2022,to assess their CSPs levels of AI adoption,technology-readiness and upcoming investment plans.The results highl

14、ight the urgent need for CSPs to address barriers to deploying AI in order to reach the level of autonomous networks.Only 6%of respondents view themselves as being at the most-advanced level of automation(zero-touch automation),which relies on AI and machine learning(ML)algorithms to manage their ne

15、tworks(see Figure 1.1).1 See Analysys Masons Telecoms capex:worldwide trends and forecasts 2018-2028.2 Telco AI refers to the use of AI-based solutions that enable the analysis of telecoms network data to address CSP business priorities such as automation,improving customer experience and operationa

16、l efficiency.Accelerating the adoption of telco AI to deliver autonomous networks|2 Analysys Mason Limited 2023 1:Executive summary Figure 1.1:CSPs current levels of automation3 The key findings of the survey are as follows.CSPs believe that AI will help them to achieve their operational and busines

17、s objectives.Improving service quality,growing revenue and improving customer experience are CSPs top-three priorities.Other priorities include energy optimisation to meet sustainability goals and operational efficiency.CSPs are deploying several AI use cases with over half running in production and

18、 the others running as either proof of concepts(PoCs)or still being explored.Network-related use cases are most common,followed by customer service-related use cases with a focus on improving operational efficiency.CSPs are recruiting expert personnel and developing strategies to ensure that they ar

19、e able to implement their telco AI initiatives.Only a small percentage of CSP respondents reported that they have invested in AI platforms.Most CSPs do not intend to develop these AI platforms themselves but plan to acquire them as a service from their cloud AI platform providers.Access to high-qual

20、ity data remains a key challenge for CSPs that want to use AI to meet their goals for autonomous networks.This challenge is impacting CSPs ability to retain AI talent and this,in turn,is affecting AI maturity.CSPs can outsource AI development and management tasks to fast-track AI use case implementa

21、tion.This demand creates opportunities for vendors that offer telecoms-specific AI solutions,but these vendors need to demonstrate that their solutions incorporate telecoms,AI,and software expertise.This perspective summarises the key findings and takeaways from our research.3 Question:“How would yo

22、u describe the current level of automation across your organisations operations?”;n=84.Accelerating the adoption of telco AI to deliver autonomous networks|3 Analysys Mason Limited 2023 2:Recommendations 2.Recommendations CSPs should quickly implement AI within networks to gain a competitive advanta

23、ge.One of CSPs key challenges to the deployment of AI use cases is their limited access to high-quality data,so they must critically evaluate their telco AI implementation strategies to circumvent this challenge.Figure 2.1 shows a two-step blueprint that CSPs can use to speed up AI use case deployme

24、nt,identify key use cases and determine implementation timelines based on data availability.This blueprint helps CSPs to prioritise investment in value-generating telco AI use cases,in line with business objectives such as operational efficiency.It also lays the foundation for pursuing other advance

25、d use cases that can drive new revenue opportunities using more-efficient networks.Figure 2.1:Two-stage blueprint for accelerating telco AI use case deployments Accelerating the adoption of telco AI to deliver autonomous networks|4 Analysys Mason Limited 2023 3:CSPs are adopting AI on their path to

26、deploying autonomous networks,but data access is the top challenge CSPs should also build the right ecosystem of vendor partners that can bring meet the needs of the telecoms domain using AI and software-as-a-service(SaaS)delivery expertise to ensure that CSPs AI investments deliver value and are co

27、st-effective.Vendors,on the other hand,should consider taking a platform-based approach to developing and deploying AI applications to support relevant use cases.The platform approach involves using common middleware resources,including data and AI capabilities to accelerate the development of AI ap

28、plications.Cloud based data and AI platforms will play a critical role because they utilise the cloud to drive agile development,delivery and operations of AI applications at scale.3.CSPs are adopting AI on their path to deploying autonomous networks,but data access is the top challenge Growing reve

29、nue,and improving service quality and customer experience are among the strategic objectives driving CSPs investment in AI CSPs believe that investment in AI will help them to meet their operational and business objectives.Analysys Masons recent survey4 shows that CSPs want to use AI to address thei

30、r top-three business priorities(see Figure 3.1).Growing revenue using AI-based tools to identify new revenue opportunities for upsell by using data about customers consumption patterns.CSPs can also increase revenue by using network intelligence to improve the services offered to customers and to de

31、velop new services(for example,using network insights to increase retailers footfall).Delivering a better customer experience using AI models to provide insights that support CSPs customer engagements,including services to provide targeted customer service/care.These are important for reducing churn

32、 and growing revenue.Improving service quality where AI algorithms automatically detect and predict issues that affect the quality of services offered to customers.Remedial steps are also prescribed or executed.As the network gets more complex,CSPs need to leverage AI tools to automate these functio

33、ns to maintain high-quality outcomes.4 Analysys Mason surveyed 84 senior CSP executives worldwide between September and November 2022 to assess each CSPs level of AI adoption,technology-readiness and upcoming investment plans.Accelerating the adoption of telco AI to deliver autonomous networks|5 Ana

34、lysys Mason Limited 2023 3:CSPs are adopting AI on their path to deploying autonomous networks,but data access is the top challenge Figure 3.1:CSPs strategic business objectives for AI5 CSPs are also using AI to address other concerns,such as reducing opex,driving operational efficiencies through au

35、tomation and sustainability.Tier-16 CSP respondents,primarily those in developed AsiaPacific,North America and Western Europe,are investing in AI to drive operational efficiencies for 5G networks.They need to grow revenue and maintain profitability to justify these investments,while also improving c

36、ustomer experience.Vodafone Italy,for example,implemented an anomaly detection solution using AI to automate network planning and optimisation functions.Tollowing this trial,Vodafone achieved increased operational efficiency of 25%to 30%as a result of the reduced time to detect and resolve issues an

37、d the associated cost savings gained from automating these workflows.CSPs energy costs are rising and 5G deployments are expected to drive energy consumption to even higher levels.This is due to the additional cell sites and active powered elements that are required to achieve low latency and to mee

38、t increased capacity demands.A more-wholistic approach to maximising energy efficiency is needed to ensure that service quality is not negatively impacted in the pursuit of reduced energy consumption.Japan-based Tier-1 CSP,KDDI,has proven in a recent trial that AI technologies can help to reduce ene

39、rgy utilisation at radio cells by up to 50%,without impairing customer experience.KDDI implemented a network energy management system that utilised ML to create models that analysed real-time demand and traffic patterns,and then automatically adjusted the amount of power consumed by RAN resources to

40、 match demand.CSPs must use AI to address these business priorities to respond to increasing competition and to the economic pressures that they currently face.The cost and revenue gains of using AI will help CSPs to improve their profitability and,more importantly,help them to enhance the level of

41、service that they deliver to their customers.5 Question:“What are your organisations business objectives for AI and automation?(Rank top-3 objectives,1=most important)?”n=84.6 Tier-1 CSPs have an annual revenue of more than USD10 billion.Accelerating the adoption of telco AI to deliver autonomous ne

42、tworks|6 Analysys Mason Limited 2023 3:CSPs are adopting AI on their path to deploying autonomous networks,but data access is the top challenge CSPs should focus on telco AI use cases that create value for their networks and their customers According to our survey,57%of CSP respondents have deployed

43、 telco AI use cases to the point of production,while 43%are either exploring or running AI use cases as PoCs.The CSP network remains a key point of investment for telco AI.Over 50%of surveyed respondents have already implemented some AI activities related to the networks in order to address the incr

44、eased complexity associated with managing higher traffic volumes and services running within the networks.The network use cases that have been most frequently deployed include network security,network design and planning use cases,with over 80%of respondents reporting that these use cases are in pro

45、duction.Customer care and experience use cases(such as AI-based chatbots,customer issue prediction and intelligent routing)have also been deployed in production by almost 60%of survey respondents.About 20%of CSP respondents are investing in new services based on data insights derived from AI.These s

46、ervices include video analytics and IoT-related services*such as smart manufacturing and autonomous driving).This gives CSPs the opportunity to provide customers(particularly enterprises)with services beyond traditional connectivity services.However,such use cases are complex to implement,given the

47、data sources and the external domain expertise that is required to support them.Figure 3.2:CSPs telco AI use case categories7 CSPs are also redefining their data and AI strategies by recruiting C-level executives(for example Chief Data Officers)to lead AI initiatives,and by modernising data infrastr

48、ucture and processes.Almost 50%of Tier-1 CSP respondents ranked having a well-defined AI strategy as a key investment that they have made to meet 7 Question:“Which of the following telco AI use case categories have you started working on?”;n=84.Accelerating the adoption of telco AI to deliver autono

49、mous networks|7 Analysys Mason Limited 2023 3:CSPs are adopting AI on their path to deploying autonomous networks,but data access is the top challenge their telco AI goals.An AI strategy provides the right structure and processes needed to accelerate AI adoption,especially given the scale of operati

50、ons at Tier-1 CSPs organisations.However,only 25%of respondents indicated that they are investing in AI platforms.These AI platforms provide the common toolsets required to create and manage AI models.As AI technologies evolve,these platforms will require continuous investment and expertise.CSPs ack

51、nowledge that they do not have the expertise to develop these platforms and will need to rely on vendor partners,such as public cloud platform vendors;CSPs can acquire this help as a service.CSPs therefore do not need to make upfront capex investments in AI platforms.CSPs main obstacle to AI adoptio

52、n is their lack of access to high-quality data CSPs still face several challenges despite ongoing investment to support telco AI.The top-ranked challenge(in terms of priority)is CSPs inability to access high-quality data,which was highlighted by 21%of the surveyed respondents(see Figure 3.3).Figure

53、3.3:Top-ranked challenges that CSPs face when developing AI use cases8 Current data infrastructures and processes are siloed,which makes it difficult for CSPs to develop AI-driven use cases.In addition,data pipelines fail to provide the data needed to support AI model development at the time require

54、d or to the quality standards expected.CSP network infrastructure and systems have proprietary interfaces,and these affect the functions of these data pipelines,especially during the data collection stages.According to our survey,data collection is a more-significant challenge than other challenges

55、for Tier-1 CSP respondents given the scale of network infrastructure that they operate.Almost 50%of Tier-1 CSPs ranked the data collection stage as the most-challenging stage of the telco AI use case development cycle.8 Question:“What are the main challenges that you are facing/or expect to face in

56、achieving telco AI/analytics goals?”;n=84.Accelerating the adoption of telco AI to deliver autonomous networks|8 Analysys Mason Limited 2023 3:CSPs are adopting AI on their path to deploying autonomous networks,but data access is the top challenge These data issues continue to impact CSPs ability to

57、 integrate AI into the network;they are also hindering CSPs AI talent retention initiatives.Executives at Tier-1 CSPs,such as Verizons Chief Data Officer and Senior Vice President,Linda Avery,reported at the Digital Transformation World 2021 that data scientists spend around 70%of their time accessi

58、ng and preparing data.Consequently,data scientists spend less time deriving insights and are less motivated to remain in such environments.Scaling AI use cases is also flagged by CSPs as a challenge,given the time and cost required to provision and manage the infrastructure and use cases as they go

59、into commercial production.These challenges affect the time to implement AI use cases.Our survey shows that the average duration to implement a telco AI use case is 67 months.At least 50%of CSPs noted that the factors influencing this timeline include the time needed to build models from scratch,as

60、well as access to the required data sets and skillsets.These issues are the result of the challenges that CSPs face when implementing AI.Figure 3.4:Factors that influence the timeline for delivering AI use cases9 CSPs should develop a clear roadmap for AI implementation,to reduce these timelines,whi

61、ch would enable them to derive more value from AI faster.Otherwise,CSPs will risk losing out on the opportunities that AI investments offer in terms of meeting their current business objectives and remaining competitive.9 Question:“What factors have influenced the timeline to deliver your AI use cas

62、es?”;n=84.Accelerating the adoption of telco AI to deliver autonomous networks|9 Analysys Mason Limited 2023 4:CSPs need a clear blueprint for deploying AI uses case and the right vendor partnerships to move to autonomous networks 4.CSPs need a clear blueprint for deploying AI uses case and the righ

63、t vendor partnerships to move to autonomous networks CSPs can move more quickly towards autonomous networks by adopting AI use cases more quickly.Factors that can help CSPs to accelerate AI adoption include creating a clearly defined blueprint for deploying AI use cases and building partnerships wit

64、h vendors that have the right skillsets.Other factors include modernising data infrastructure and using the clouds scalability and elasticity to fast-track deployments.This section of the paper discusses defining a blueprint for implementing AI use cases and building the right vendor partnerships.CS

65、Ps should adopt an AI use case implementation blueprint,which should be driven by the ease of access to data and the time that it takes to derive value Given the competing priorities that CSPs have and the challenges that they expect AI to address a well-defined blueprint should help CSPs to develop

66、 a strategy to deploying AI use cases.This blueprint should consider which use cases to implement,and the timescales required to develop and deploy them.Figure 4.1 shows the factors that will help CSPs to identify and prioritise AI use case investment.Accelerating the adoption of telco AI to deliver

67、 autonomous networks|10 Analysys Mason Limited 2023 4:CSPs need a clear blueprint for deploying AI uses case and the right vendor partnerships to move to autonomous networks Figure 4.1:Factors that CSPs should consider when selecting which AI use cases to deploy CSPs should start by identifying AI u

68、se cases that can address their top business priorities.This step is critical for ensuring that CSPs target value-generating use cases.Once use cases are selected,as well as the data sets to help them deploy the use cases,the relevant data sources(for example network equipment and OSS systems)should

69、 be identified.CSPs need to determine how quickly data can be accessed from these sources,as well as the quality of the data that they generate.Inaccessible and poor-quality data sets make it difficult and time-consuming to develop use cases.Finally,CSPs should determine the expertise and skillsets

70、required to implement use cases.Telco domain and AI expertise will be relevant as they ensure that use cases deliver accurate and relevant insights.CSPs can move on to define the timescales for developing telco AI use cases.CSPs should determine these timescales based on how quickly the use cases ca

71、n deliver value,on the effort that is required to implement them.Value should be measured based on each use cases ability to help CSPs to achieve their top business priorities.These priorities are often measured using key performing indicators(KPIs)such as revenue growth Accelerating the adoption of

72、 telco AI to deliver autonomous networks|11 Analysys Mason Limited 2023 4:CSPs need a clear blueprint for deploying AI uses case and the right vendor partnerships to move to autonomous networks and savings in cost and time.Ease of implementation should be determined based on access to data and exper

73、tise and the complexity of the algorithms involved in creating the use case.Following this assessment,CSPs should be able to categorise the use cases as follows.Use cases that can be implemented in the short term.These use cases address immediate concerns about internal network operations,such as im

74、proving performance in network infrastructure management and operations,and energy optimisation.They are relatively easy to deploy because the data required to train the models can be derived from internal systems within the same network organisation.The approvals to access the data are provided by

75、internal CSP network stakeholders.Examples include automated incident detection,alarm reduction,automated health checks of network infrastructure,energy optimisation and smart capex allocation.Use cases that can be implemented in the mid-term:The data sets required to deliver these use cases come fr

76、om multiple network(that is access,core,and transport)and non-network related teams.Examples include service quality monitoring and end-to-end service orchestration.End-to-end service orchestration,for example,will require creating AI/ML models that can detect issues occurring within multiple networ

77、k domains and in response,can initiate remedial actions via the orchestrator.Use cases that can be implemented in the long term:These use cases are often driven by the need to capture new revenue.They require access to third-party data sets and so could be more complex to achieve and may require mor

78、e investment.For example,the model development and deployment processes require more time than the internal use cases and external industry expertise.Examples of these use cases include video and IoT device analytics.When using AI for video analytics,large data storage infrastructure will be require

79、d.It can take a long time to assemble the right approval to access data,pre-processing the unstructured video data sets and developing models that combine network and video-related data.Consequently,the time to generate value for these use cases is longer than the internal operational use cases and

80、should be considered as a long-term use case.Several benefits can be derived from adopting this blueprint approach to implementing telco AI use cases.By first deploying use cases that focus on optimising the operations of the network infrastructure,CSPs are able to position themselves to support the

81、 needs of customers,especially those in the enterprise market.The implementation of domain-or infrastructure-specific use cases first also lays the foundation for developing cross-domain-related use cases such as end-to-end service orchestration.Additional knowledge and expertise can also be gained

82、from deploying the short-to mid-term use cases,which can be leveraged to drive long-term use cases.CSPs should build partnerships with vendors that have combined expertise in telecoms,AI and software According to our survey,45%of surveyed CSP respondents plan to outsource the development of their AI

83、 use cases.By outsourcing AI development,CSPs can fast-track their AI projects as they get access to AI expertise;this helps them to avoid the delays that they would otherwise face if trying to attract and hire staff with these skills.However,in addition to having a third-party develop the AI use ca

84、ses,core responsibilities such as data and model management will need to be addressed.Consequently,CSPs will require a partner that has the relevant expertise to meet their expectations.In our research,surveyed CSP respondents were asked to rank their vendor selection criteria(see Figure 4.2 for a s

85、ummary of their responses).Accelerating the adoption of telco AI to deliver autonomous networks|12 Analysys Mason Limited 2023 4:CSPs need a clear blueprint for deploying AI uses case and the right vendor partnerships to move to autonomous networks Figure 4.2:CSPs top vendor selection criteria for A

86、I solutions10 23%of the surveyed CSP respondents ranked vendors combined telco and AI expertise and managed services capabilities more highly than other selection criteria.Vendors with deep telco domain knowledge understand CSPs networks and operations and have a clear view of the pain points that c

87、an be addressed with AI.These vendors can also bring a wealth of experience to designing,deploying,and managing networks and associated operations,to validate the insights obtained from AI models before they are deployed into the CSP environment.In addition to having telco and AI expertise,15%of sur

88、veyed CSPs want to partner with vendors that can offer AI solutions using the software-as-a-service(SaaS)delivery model.CSPs are finding SaaS-based solutions easy to use(relative to the complexity of the problem to be addressed)but also agile and scalable in response to changes occurring in the CSP

89、environment.While SaaS-based offerings are attractive to CSPs,they will involve vendors meeting CSPs stringent security requirements,especially in relation to where and how CSPs data is stored and managed.Current data sovereignty regulations will place more pressure on vendors that offer SaaS-based

90、AI solutions to align with these requirements.Vendors that offer AI solutions to CSPs should also consider leveraging cloud platform services to deliver these solutions.CSPs are subscribing to data and AI cloud-based platform services as part of their runtime environments because they are scalable a

91、nd offer CSPs the opportunity to gain faster access to the innovative data and AI services from hyperscalers and other cloud data and AI solution providers.In our research,70%of CSPs would expect their vendor partners to either become direct consumers of these cloud-based data and AI platform servic

92、es or become providers of application platforms with capabilities to offer telco-specific platform services that run in the cloud.CSPs expect their vendor partners to take a platform-based approach to delivering their AI use cases.This approach creates opportunities for CSPs to reuse platform capabi

93、lities to accelerate AI development.It also enables vendors that offer AI-enabled solutions to develop AI use cases at a faster pace as reusable assets present within the platform and that can be extended to support new use cases.Vendors can also expose their 10 Question:“What factors would you cons

94、ider when selecting a vendor partner to support your telco Ai strategies?”;n=84.Accelerating the adoption of telco AI to deliver autonomous networks|13 Analysys Mason Limited 2023 5:Conclusion platforms to CSPs(those that have the software expertise)to develop use cases that align with their specifi

95、c needs.For those vendors that want to target telecoms operators with AI-enabled solutions,enhancing their telco expertise,SaaS delivery capabilities and taking advantage of cloud-based AI and data platform services will be critical to building their competitive advantage.5.Conclusion Cellular data

96、traffic is expected to more than triple between 2021 and 2026 to 2571ZB,11 so CSPs must transition to more-autonomous operations in order to manage networks more efficiently and to meet their main business priorities.Telco AI will be a core capability required to make this change.However,for most CS

97、Ps,accessing high-quality data remains the main obstacle to deploying telco AI within their networks.Data infrastructure and processes operate in silos,and data pipelines fail to deliver data in the timeframe and at the quality levels required to support the functions that help to integrate AI capab

98、ilities into network operational workflows.Analysys Masons survey highlights the critical role that telco AI will play as CSPs aim for more-autonomous operations.It also shows CSPs readiness to implement this technology and the key factors that can help CSPs to accelerate the deployment of telco AI.

99、87%of CSPs have started to implement AI,either as PoCs or into production,leaving those who are yet to begin their telco AI activities at a disadvantage.CSPs will need to define a clear roadmap for implementing AI in order to meet their ambitions for autonomous networks.CSPs must identify and priori

100、tise which use cases to deploy,and they must set out the deployment timelines for these use cases.These steps need to be performed by taking the value of the use cases into consideration,as well the ease of access to that data required that is required to deliver them.CSPs should also create partner

101、ships with vendors that have telco domain,AI and software expertise in order to fast-track AI implementation.11 For more information,see Analysys Masons DataHub.Accelerating the adoption of telco AI to deliver autonomous networks|14 Analysys Mason Limited 2023 About the author 6.About the author Ada

102、ora Okeleke(Principal Analyst)leads Analysys Masons Data,AI and Development Platforms research programme.Her research focuses on service providers adoption and use of data management,artificial intelligence,analytics and development tools to support the digital transformation of network,customer and

103、 other business operations.Adaora tracks vendor strategies for the telecoms industry to understand how they are evolving their product portfolios to include data,AI and development capabilities.She also provides key industry insights to operators and vendors on strategies for adopting these technolo

104、gies.12 12 Analysys Mason Limited.Registered in England and Wales with company number 05177472.Registered office:North West Wing Bush House,Aldwych,London,England,WC2B 4PJ.We have used reasonable care and skill to prepare this publication and are not responsible for any errors or omissions,or for th

105、e results obtained from the use of this publication.The opinions expressed are those of the authors only.All information is provided“as is”,with no guarantee of completeness or accuracy,and without warranty of any kind,express or implied,including,but not limited to warranties of performance,merchan

106、tability and fitness for a particular purpose.In no event will we be liable to you or any third party for any decision made or action taken in reliance on the information,including but not limited to investment decisions,or for any loss(including consequential,special or similar losses),even if advi

107、sed of the possibility of such losses.We reserve the rights to all intellectual property in this publication.This publication,or any part of it,may not be reproduced,redistributed or republished without our prior written consent,nor may any reference be made to Analysys Mason in a regulatory statement or prospectus on the basis of this publication without our prior written consent.Analysys Mason Limited and/or its group companies 2023.