《易观梅森:2023年端到端自动化:机遇、挑战和部署状态洞察报告(英文版)(26页).pdf》由会员分享,可在线阅读,更多相关《易观梅森:2023年端到端自动化:机遇、挑战和部署状态洞察报告(英文版)(26页).pdf(26页珍藏版)》请在三个皮匠报告上搜索。
1、 Perspective End-to-end automation:opportunities,challenges and the state of deployment April 2023 Justin van der Lande,Michelle Lam and Hansang(Andy)HeEnd-to-end automation:opportunities,challenges and the state of deployment|i Analysys Mason Limited 2023 Contents Contents 1.Executive summary 1 2.A
2、utomation is critical to ensuring success in the 5G era 4 Key drivers of automation 5 Automation will help CSPs to tackle the complexities of 5G 7 3.What is the state of CSPs closed-loop automation deployments today?8 Who is in charge?8 Level of automation 8 Challenges 10 Financial impact 11 Budget
3、allocation 11 Technology and vendor selection 12 4.Measuring the success of automation 12 KPI categories 12 KPIs for end-to-end automation processes 13 Expected improvement across all lifecycle processes 21 5.Recommendations 21 6.About the authors 23 List of figures Figure 1.1:Job titles of the surv
4、ey respondents,worldwide,2022.2 Figure 1.2:Responsibilities held by the survey respondents,worldwide,2022.2 Figure 2.1:The role of automation in network and service orchestration.5 Figure 2.2:Factors driving CSPs automation initiatives,worldwide,2022.6 Figure 2.3 Top challenges that CSPs are using a
5、utomation to overcome,worldwide,2022.7 Figure 3.1:CSPs current automation levels for their assurance and orchestration processes,worldwide,2022.9 Figure 3.2:CSPs future plans for automation,worldwide,2022.9 Figure 3.3:CSPs main challenges to achieving their automation goals,worldwide,2022.10 Figure
6、3.4:Ranking of network and service lifecycle processes in terms of the financial impact of the deployment of automation solutions,worldwide,2022.11 Figure 3.5:Average budget per automation project,by region,worldwide,2022.12 Figure 4.1:Improvements in the KPIs that CSPs rank as one of their top two
7、KPIs for the network plan to build virtual infrastructure process since 2020,worldwide,2022.14 Figure 4.2:Expected improvements in the KPIs that CSPs rank as one of their top two KPIs for the network plan to build virtual infrastructure process in the next 24 months,worldwide,2022.15 Figure 4.3:Impr
8、ovements in the KPIs that CSPs rank as one of their top two KPIs for the service design to build process since 2020,worldwide,2022.16 Figure 4.4:Expected improvements in the KPIs that CSPs rank as one of their top two KPIs for the service design to build process in the next 24 months,worldwide,2022.
9、16 End-to-end automation:opportunities,challenges and the state of deployment|ii Analysys Mason Limited 2023 Contents Figure 4.5:Improvements in the KPIs that CSPs rank as one of their top two KPIs for the order to activate process since 2020,worldwide,2022.17 Figure 4.6:Expected improvements in the
10、 KPIs that CSPs rank as one of their top two KPIs for the order to activate process in the next 24 months,worldwide,2022.18 Figure 4.7:Improvements in the KPIs that CSPs rank as one of their top two KPIs for the problem to resolution process since 2020,worldwide,2022.19 Figure 4.8:Expected improveme
11、nts in the KPIs that CSPs rank as one of their top two KPIs for the problem to resolution process in the next 24 months,worldwide,2022.19 Figure 4.9:Improvements in the KPIs that CSPs rank as one of their top two KPIs for the termination to confirmation process since 2020,worldwide,2022.20 Figure 4.
12、10:Expected improvements in the KPIs that CSPs rank as one of their top two KPIs for the termination to confirmation process in the next 24 months,worldwide,2022.21 This perspective was commissioned by Hewlett Packard Enterprise.Usage is subject to the terms and conditions in our copyright notice.An
13、alysys Mason does not endorse any of the vendors products or services.End-to-end automation:opportunities,challenges and the state of deployment|1 Analysys Mason Limited 2023 1:Executive summary 1.Executive summary Automation has been a strategic focus for communications service providers(CSPs)for m
14、any years in order to improve operational efficiency,offer higher-quality services and reduce the time to market for new services.5G deployments have acted as a catalyst for the acceleration of automation due to their requirements for cloud-native and microservices-based development,software-defined
15、 and highly-virtualised network architecture and new services that support the Industry 4.0 era and other advanced use cases.Extreme network automation is also required when adopting open and disaggregated RAN architecture to reduce the cost of integration across multi-vendor environments using stan
16、dardised,open and interoperable interfaces.All of these factors mean that CSPs must increase their levels of automation to be able to realise new revenue opportunities such as network slicing,software-defined networking(SDN)and cost savings using network function virtualisation(NFV).Analysys Mason c
17、onducted a study in October 2022 to assess the current state of CSPs automation journeys.This paper provides in-depth analysis of the results of the study,which comprised a survey of 70 Tier-1 and Tier-2 CSPs worldwide(Figure 1.1 and Figure 1.2),1 interviews with a subset of 10 CSPs and Analysys Mas
18、ons existing network automation and orchestration research.The study focused on areas relating to CSPs automation initiatives such as the challenges,benefits,financial impacts,use case priorities,deployment timelines and key performance indicators(KPIs)for multiple end-to-end processes including the
19、 process from network planning to building virtual infrastructure,that from designing a service to building it and the processes from ordering to activating,from problem to resolution and from termination to confirmation.1 Respondents specialities are the core network(57%),fixed access(56%),the tran
20、sport network(34%),the RAN(21%)and edge networks(16%).Their products and services are tailored to the B2B(36%),B2C(9%)and both the B2B and the B2C(56%)markets.End-to-end automation:opportunities,challenges and the state of deployment|2 Analysys Mason Limited 2023 1:Executive summary Figure 1.1:Job t
21、itles of the survey respondents,worldwide,20222 Figure 1.2:Responsibilities held by the survey respondents,worldwide,20223 The key findings from our survey are as follows.The four main drivers of automation are cost avoidance(53%),shorter time to launch new services(43%),new revenue creation(41%)and
22、 improving operational accuracy(41%).This shows that CSPs are currently 2 Question:“What is your job title?”;n=70.3 Question:”Do you have any responsibility for the following areas?”;n=70.End-to-end automation:opportunities,challenges and the state of deployment|3 Analysys Mason Limited 2023 1:Execu
23、tive summary placing a stronger emphasis on using automation to address financial and operational challenges than to enhance the customer experience,because improving the Net Promoter Score(NPS)is ranked as the fifth most important driver(39%).The biggest challenges that CSPs are using automation to
24、 tackle are overcoming the complexities of a microservices-based 5G core to deploy 5G standalone(34%),simplifying workloads to accelerate 5G roll-out(24%)and eliminating the high costs associated with deploying Open RAN(16%).5G standalone has complex network requirements(in terms of connection densi
25、ty,rate,latency and reliability),which requires a high level of workload automation.Vendors must re-engineer their portfolios around microservices-based infrastructure,containerised workloads and orchestration solutions to support new 5G use cases.CSPs and vendors must work closely together to deplo
26、y pilot vRAN and ORAN sites,and should develop advanced orchestration and automation solutions for the virtualisation of the RAN.A lack of technology maturity(46%),change management(46%)and a lack of the right skillsets(39%)were cited as the top three challenges faced by CSPs on their automation jou
27、rneys.CSPs must invest in the research and development of new automation technologies(SDN/NFV,network slicing and edge compute)and should consider organisational changes to enable them to embrace new industry standards and architecture frameworks.They should also reskill their workforces to ensure t
28、hat they have the requisite software engineering skills to advance in their automation journeys.70%of respondents agreed that it is beneficial if assurance and orchestration processes are integrated by the same vendor,while 11%of CSPs disagreed.19%are still unsure.We predict that most CSPs will cont
29、inue to scale up their single-vendor virtualised network function(VNF)orchestrators,and vendors must ensure that they continue to advance their automation and orchestration solutions for NFV-based services.61%of the surveyed CSPs expect that a vendor-agnostic,zero-touch automation solution with serv
30、ice modelling will help to resolve most of their automation-related challenges.CSPs are looking to adopt new operating models that are built on a multi-vendor strategy and that enable greater agility and flexibility through the use of best-of-breed products and applications.An automation solution th
31、at enables vendor-agnostic service lifecycle management and can orchestrate all the exposed network applications and resources to provision,manage and control services on demand will help to accelerate CSPs automation goals.The process for which deploying automation has the greatest financial impact
32、(ranked in the top two)is service design to build(53%),followed by network plan to build virtual infrastructure(48%),problem to resolution(47%),order to activate(33%)and termination to confirmation(19%).Vendors must ensure that they continue to improve their automation credentials by offering a serv
33、ice and network automation and orchestration solution that covers all aspects of service lifecycle management(testing,onboarding,monitoring and operation)and network planning and design.70%of CSPs have already started to automate their assurance and orchestration functions.20%have plans to begin the
34、ir automation journeys in the next 12 months and the remainder have no plans for automation.CSPs that have already begun their automation journeys have a long way to go until they achieve fully autonomous networks with closed-loop operations.There are likely to be many partially automated processes
35、across the entire service lifecycle whereby manual processes are supported but not entirely replaced.End-to-end automation:opportunities,challenges and the state of deployment|4 Analysys Mason Limited 2023 2:Automation is critical to ensuring success in the 5G era 2.Automation is critical to ensurin
36、g success in the 5G era Next-generation 5G networks are causing existing operations models to become unsustainable because management tools are not able to support new technologies,new service types,the higher cadence of network and service changes and the demands for rapid self-service delivery.Est
37、ablished systems will therefore require manual intervention from technicians unless automations are implemented.The ability to automate both the network and services will enable CSPs to rapidly launch and support next-generation service offerings at scale,thereby providing the foundation for the dig
38、ital transformation of enterprises across various industries.As such,CSPs are working towards enabling a high degree of automation and zero-touch management throughout all phases of the network and service lifecycle.Network plan to build virtual infrastructure.The network plan and build stage covers
39、 operational processes such as project management,surveying,designing,planning,estimation,partner and supplier management and implementation.Service design to build.This process deals with all the activities related to the launch of a new product/service and the retirement of an old product/service.
40、An idea for a new product or service must be transformed into a ready-to-sell entity with process flows,integrations,tariffs and assurance.Order to activate.This process takes a customer order and implements it in the network and other infrastructure.It also supports service testing prior to going l
41、ive and informs customers of any change in the status of their order.Problem to resolution.This process deals with activities that allow a problem raised by a customer or technical support team to be solved.It also addresses problems that are proactively identified internally,even before the custome
42、r or service is affected.Termination to confirmation.This process deals with all the activities related to the execution of a customers termination request.It also involves activities that can prevent customers from terminating.Network automation eliminates the manual tasks that are associated with
43、the configuration,management,operations and lifecycle of network services and enables autonomous operations.Service automation involves the use of continuous integration/continuous delivery(CI/CD)tools to automate and monitor the lifecycle of apps,from the integration and testing phases to delivery
44、and deployment.Figure 2.1 describes how the role of automation in network and service orchestration supports CSPs in deploying end-to-end automation to deliver differentiated 5G services in an agile and scalable manner,reduce the time to revenue and improve operational efficiency.End-to-end automati
45、on:opportunities,challenges and the state of deployment|5 Analysys Mason Limited 2023 2:Automation is critical to ensuring success in the 5G era Figure 2.1:The role of automation in network and service orchestration Key drivers of automation CSPs success in the 5G era will be defined by their abilit
46、y to launch and operationalise new digital services in an agile and cost-effective manner.The existing telecoms operational model is not fit for purpose for the 5G era because business processes such as network design and planning,network roll-out and provisioning,assurance,operations and maintenanc
47、e are highly disjointed and often require manual interventions and manual inter-departmental handovers.This results in high opex and limited service agility.As such,the key drivers of automation are cost avoidance,new revenue creation and an improvement in operational accuracy(Figure 2.2).End-to-end
48、 automation:opportunities,challenges and the state of deployment|6 Analysys Mason Limited 2023 2:Automation is critical to ensuring success in the 5G era Figure 2.2:Factors driving CSPs automation initiatives,worldwide,20224 Cost avoidance CSPs are under pressure to significantly reduce their operat
49、ional costs because traditional methods of operating the network and service lifecycle are costly and inefficient.The shift to 5G standalone(SA)networks,which involves a more-radical change of architecture than is needed for 5G non-standalone(NSA)networks,carries a huge potential cost for CSPs,many
50、of which are not yet clear about how to monetise these networks.The desire to avoid additional costs is reflected in our survey results;more than 56%of respondents identified cost avoidance as the biggest commercial driver of automation.Cost avoidance is also the most important driver for CSPs in No
51、rth America.New revenue creation The emergence of advanced 5G network capabilities(network slicing,low latency and ultra-reliable connectivity)creates the potential for new revenue opportunities in the enterprise segment.However,a high degree of automation is required to scale up the monetisation an
52、d delivery of these new digital experiences and advanced B2B services to digital-native customers.Indeed,42%of respondents cited new revenue creation as a key driver of automation.New revenue creation is also the most important driver for CSPs in developed AsiaPacific(DVAP).Improvement in operationa
53、l accuracy CSPs are deploying closed-loop automation to reduce the number of errors introduced during service fulfilment processes.This is particularly important for service modifications,which typically require network operations staff to manually delete the service and provision a new version with
54、 a modified configuration.The remote and automated monitoring,troubleshooting,configuration and optimisation of network resources and service lifecycles will enable greater efficiency in network operations and ensure a high quality of service(QoS).41%of respondents in our survey cited improving oper
55、ational accuracy as a key driver of automation.4 Question:“What is driving your automation initiatives?”;n=70.End-to-end automation:opportunities,challenges and the state of deployment|7 Analysys Mason Limited 2023 2:Automation is critical to ensuring success in the 5G era Automation will help CSPs
56、to tackle the complexities of 5G 5G network deployments require a new generation of orchestration technologies and CSPs will need new automation capabilities to overcome the complexities of deploying,operating and configuring 5G networks to support new use cases.The top three challenges that CSPs ar
57、e using automation to address are overcoming the complexities of a microservices-based 5G core to deploy 5G SA,simplifying the exponential growth in workloads associated with 5G roll-outs and eliminating the high costs associated with deploying Open RAN(Figure 2.3).CSPs are evaluating new approaches
58、 to network and service automation to ensure that they remain competitive.They are also using automation to reduce the need for manual interventions so that they can both respond more quickly to new service orders,modifications and network issues and provide real-time experiences.Figure 2.3 Top chal
59、lenges that CSPs are using automation to overcome,worldwide,20225 5G SA The number of 5G mobile network deployments grew considerably in 2022,and many more launches are scheduled for 2023.Some of these are expected to use 5G SA architecture.CSPs are increasingly keen to showcase their credentials in
60、 terms of both the coverage and the architecture of the network that they are deploying.They are accelerating their shift to SA networks(which support the full range of 5G capabilities)as the need to deliver a fully 5G experience to end users becomes more compelling.However,rearchitecting the 5G cor
61、e to be fully cloud-native and independent of 4G will increase the complexity of the 5G architecture,so zero-touch end-to-end network automation needs to be employed in order to scale 5G SA deployments.Open RAN CSPs that wish to deploy the Open RAN can benefit from new control and management archite
62、cture to enable the fine-grained programmatic control of RAN resources based on evolving network conditions and to deliver a range of use cases with various requirements for QoS,network performance,service-level agreements(SLAs)5 Question:“Which are the top 3 challenges you are trying to tackle usin
63、g automation?”;n=70.End-to-end automation:opportunities,challenges and the state of deployment|8 Analysys Mason Limited 2023 3:What is the state of CSPs closed-loop automation deployments today?and latency.The RAN intelligent controller(RIC)gives CSPs the ability to control and manage RAN resources
64、at a more granular level and enables RAN automation that is more intelligent and use-case-driven.Network slicing 5G enables new technologies such as network slicing.Network slicing in turn enables networks to be partitioned into multiple isolated,virtual network slices with unique characteristics(in
65、 terms of quality of service,latency and bandwidth)that are specific to the requirements of the service supported.5G network slicing is a key enabler for CSPs to efficiently manage network resources and offer differentiated SLA-based services to improve service agility.5G network slicing requires ad
66、vanced orchestration solutions that can work across domains and vendors,as well as automation solutions that can create,deploy and manage slices at scale.“The motivation here is to look from the virtualisation and cloudification perspective.The relationship is very complex once you know CNFs and VNF
67、s looks and perceptions.So,how we can manage this complex architecture of CNFs and VNFs?The solution is automatic orchestration.We have an opportunity to progress considerably and automate the operations of CNFs and VNFs.Senior Product Manager/Lead Design&Architecture,Tier-1 CSP,UK 3.What is the sta
68、te of CSPs closed-loop automation deployments today?Who is in charge?The majority of CSPs automation projects are currently led by CIO teams(36%),followed by CTO(33%)and CMO(30%)teams.Respondents to our survey reported that automation is more of an IT issue than a networking operations issue.Indeed,
69、77%of respondents reported that their IT department is the most involved in their automation initiatives.Level of automation 46%of respondents have automated 25%of their assurance and orchestration functions(Figure 3.1),but almost all plan to increase their level of automation(Figure 3.2).End-to-end
70、 automation:opportunities,challenges and the state of deployment|9 Analysys Mason Limited 2023 3:What is the state of CSPs closed-loop automation deployments today?Figure 3.1:CSPs current automation levels for their assurance and orchestration processes,worldwide,20226 Figure 3.2:CSPs future plans f
71、or automation,worldwide,20227 6 Question:“How far have you automated your assurance and orchestration functions(closed-loop orchestration)?”;n=70.7 Question:“Do you plan to increase further the degree of automation for your assurance and orchestration functions(closed-loop orchestration)?”;n=70.End-
72、to-end automation:opportunities,challenges and the state of deployment|10 Analysys Mason Limited 2023 3:What is the state of CSPs closed-loop automation deployments today?“It is going to be on the top at the group CTO level to update the strategic plan for CSPs automation journeys,and then each enti
73、ty will have to take that into consideration;we are still on trying to catch up,for example,on the tools in the CI/CD chain and what were going to use to propose or suggest and offer commonalities.”Director of IT/Network Strategy and Transformation for the Middle East and Africa,Tier-1 CSP,Europe Ch
74、allenges CSPs still face challenges to deploying end-to-end network automation due to the difficulties in orchestrating network functions.Internal issues,both cultural and technological,are the key barriers to the widespread adoption of automation solutions.Respondents cited a lack of technology mat
75、urity(46%),change management(46%)and a lack of the right skillsets(39%)as the top three challenges that they face on their automation journeys(Figure 3.3).Figure 3.3:CSPs main challenges to achieving their automation goals,worldwide,20228 Organisational resistance to change is the most significant b
76、arrier to deploying automation.This implies that respondents feel that their organisations will resist changes despite them being beneficial.CSPs also lack the technology maturity and in-house technical expertise and skillsets required to implement automation solutions that comply with the latest so
77、ftware engineering paradigms such as container-and microservices-based development for cloud deployment and agile principles such as DevOps and CI/CD pipelines.Furthermore,CSPs current IT systems and network infrastructure are highly customised and built upon silos,and processes such as network desi
78、gn,planning,fulfilment,assurance and operations are highly disjointed,thereby often requiring manual interventions and manual interdepartmental handovers.8 Question:“What are the main challenges you expect to face/are facing in achieving your automation end goal?”;n=70.End-to-end automation:opportun
79、ities,challenges and the state of deployment|11 Analysys Mason Limited 2023 3:What is the state of CSPs closed-loop automation deployments today?Financial impact Respondents were asked to rank network and service lifecycle processes in terms of the financial impact of the deployment of automation so
80、lutions(Figure 3.4).Service design to build was ranked in the top two by 53%of respondents;this is the largest share for any process.This was followed by network plan to build virtual infrastructure(48%),problem to resolution(47%),order to activate(33%)and termination to confirmation(19%).Figure 3.4
81、:Ranking of network and service lifecycle processes in terms of the financial impact of the deployment of automation solutions,worldwide,20229 “Making a payment depends on the mediation role.The mediation role is about USD1 million.Where the RAN5 got to a different level,we are talking about USD300
82、million last year,because the department of the whole cell size infrastructure,and the core network is about USD200 million plus programmes and the BSS is about USD50 million.Even some of the programs,for example,some of the OSS tools,had no network,and they were as small as,for example,USD800 000 o
83、r less than USD1 million.The automation mandate was spread across various programs,then how much they can invest all of that for the automation part will be quite lenient.In about 2 years,we should save about USD40 million.”Senior Product Manager/Lead Design&Architecture,Tier-1 CSP,UK Budget allocat
84、ion Most of the CSPs surveyed reported that automation projects account for 2130%of their operations budget,and that they had carried out 3140 automation projects in the past 3 years.CSPs tend to allocate most of their budgets to in-house teams or managed services providers when implementing automat
85、ion solutions.Indeed,35%of CSPs automation budgets are allocated to in-house teams,followed by managed service providers 9 Question:“Please rank the following processes(please refer to the process diagram)in terms of greatest financial impacts brought about by the deployment of automation solutions
86、for your business”;n=70.End-to-end automation:opportunities,challenges and the state of deployment|12 Analysys Mason Limited 2023 4:Measuring the success of automation (18%),prime system integrators(16%),key network equipment providers(14%),OSS vendors(12%)and others(5%).Our survey results also show
87、 that CSPs from the Middle East have the highest average budget per automation project,followed by those in Latin America and Western Europe(Figure 3.5).Figure 3.5:Average budget per automation project,by region,worldwide,202210 Technology and vendor selection CSPs have an overwhelming preference to
88、 use components from a single vendor.Indeed,70%of respondents agreed that it is beneficial if assurance and orchestration processes are integrated by the same vendor,11%disagreed and 19%are still unsure.Most CSPs are currently deploying single-vendor VNF orchestrators.Network disaggregation will ena
89、ble CSPs to move away from the traditional operating model and take a multi-vendor strategy that offers the ability to select best-of-breed products and applications.61%of CSPs believe that a vendor-agnostic,zero-touch automation solution with service modelling will resolve most of their automation-
90、related challenges.4.Measuring the success of automation KPI categories CSPs are using various KPIs to monitor their network performance and QoS as a result of deploying automation solutions.These KPIs can be split into the following categories.10 Question:“What region is your organisations headquar
91、ters located in?”;n=70.End-to-end automation:opportunities,challenges and the state of deployment|13 Analysys Mason Limited 2023 4:Measuring the success of automation Time taken.Using automation to reduce the time taken to complete a process will help to decrease costs due to the reduction in the nu
92、mber of time-consuming,manual processes.Further advantages of lowering the time taken to complete a process include a decrease in backlogs(including for orders and payments)and improvements to customer satisfaction.Reliability.This refers to the ability of a system or component to perform its requir
93、ed functions under stated conditions for a specific period of time.Increasing reliability will help to reduce the costs associated with maintenance and manual intervention.Capacity.This is a measure of the actual output of a process compared to its potential maximum output.Automation will enable pro
94、cesses to be executed at scale.For example,whole network equipment upgrades could be carried out in one go.Integrity.This involves measuring the quality of the network,throughput and latency,making improvements and ensuring that SLAs are met.“For KPIs,we have about five categories.Accessibility,reli
95、ability and sustainability are the major ones for the service department.The other two categories are accountability for postal service deployment and functions.Service provisioning mainly involves accountability,accessibility,reliability and sustainability.For the fault automation part,you are main
96、ly concerned about how to deploy and set up those KPIs automatically,largely for zero-touch provisioning.Integrating APIs and VNFs based on different cloud platforms is a major challenge because each vendor has its own interfaces,the APIs and their own VNF-like container design.They need to have goo
97、d interoperability on different platforms,which is another challenge.”Principal RAN Engineer,Tier-1 CSP,USA KPIs for end-to-end automation processes Network plan to build virtual infrastructure Network lifecycle management involves the provisioning and upgrade of network equipment and can be highly
98、automated.Indeed,many CSPs are already in the process of automating such processes;basic use cases include network site planning,network resource optimisation,data correlation and network troubleshooting and configuration.The time to configure edge compute infrastructure is the most commonly used KP
99、I to measure the success of automation during the network plan to build virtual infrastructure process.39%of the respondents that ranked this KPI as one of their two top KPIs for the network plan to build virtual infrastructure process have seen an improvement of 2136%since 2020(Figure 4.1),and 44%e
100、xpect to see a 3750%improvement in the next 24 months(Figure 4.2).The number of manual interventions required to onboard new VNFs and CNFs was the second-most-popular KPI to measure the success of automation in the network plan to build virtual infrastructure process.42%of the respondents that ranke
101、d this KPI as one of their two top KPIs for the network plan to build virtual infrastructure process have seen an improvement of 2136%since 2020,and 25%have seen improvements of 51%or more(Figure 4.1).58%expect to see a 3750%improvement in the next 24 months(Figure 4.2).It is noteworthy that the onb
102、oarding of xNFs still poses a major challenge more than 10 years since the advent of NFV.End-to-end automation:opportunities,challenges and the state of deployment|14 Analysys Mason Limited 2023 4:Measuring the success of automation Figure 4.1:Improvements in the KPIs that CSPs rank as one of their
103、top two KPIs for the network plan to build virtual infrastructure process since 2020,worldwide,202211 11 Question:“What KPIs are you using to measure the success of automation in the network plan to build virtual infrastructure process?For your top two KPIs,what result have you observed since 2020?”
104、;n=34.Note that respondents did not spell out whether the edge infrastructure was referring to RAN or vCPE/SDN.Given the current pace of the market development,it is likely that the current focus is on RAN/SDN and in the next 24 months the focus could shift to vCPE.End-to-end automation:opportunitie
105、s,challenges and the state of deployment|15 Analysys Mason Limited 2023 4:Measuring the success of automation Figure 4.2:Expected improvements in the KPIs that CSPs rank as one of their top two KPIs for the network plan to build virtual infrastructure process in the next 24 months,worldwide,202212 S
106、ervice design to build The time taken to deliver new services is the most commonly used KPI to measure of the success of automating the service design to build process.Specifically,the average time to market(TTM)for B2B products and services is the most popular KPI for this process.CSPs have seen la
107、rge improvements in this KPI since 2020(Figure 4.3),and further improvements are expected over the next 24 months(Figure 4.4).The average TTM for B2C products and services is the second most-popular KPI,but 14%of CSPs that ranked this as one of their top two KPIs for the service design to build proc
108、ess have not seen any improvement in performance since 2020(Figure 4.3).Nonetheless,more than half of CSPs expect improvements of 3750%for this KPI in the next 24 months(Figure 4.4).12 Question:“What KPIs are you using to measure the success of automation in the network plan to build virtual infrast
109、ructure process?For your top two KPIs,what is your desired goal for the next 24 months?”;n=34.Note that respondents did not spell out whether the edge infrastructure was referring to RAN or vCPE/SDN.Given the current pace of the market development,it is likely that the current focus is on RAN/SDN an
110、d in the next 24 months the focus could shift to vCPE.End-to-end automation:opportunities,challenges and the state of deployment|16 Analysys Mason Limited 2023 4:Measuring the success of automation Figure 4.3:Improvements in the KPIs that CSPs rank as one of their top two KPIs for the service design
111、 to build process since 2020,worldwide,202213 Figure 4.4:Expected improvements in the KPIs that CSPs rank as one of their top two KPIs for the service design to build process in the next 24 months,worldwide,202214 13 Question:“What KPIs are you using to measure the success of automation in the servi
112、ce design to build process?For your top two KPIs,what result have you observed since 2020?”;n=37.14 Question:“What KPIs are you using to measure the success of automation in the service design to build process?For your top two KPIs,what is your desired goal for the next 24 months?”;n=37.End-to-end a
113、utomation:opportunities,challenges and the state of deployment|17 Analysys Mason Limited 2023 4:Measuring the success of automation Order to activate The average handling time(AHT)for B2C products/services and the difference between the scheduled and actual provisioning time of service instances are
114、 the most commonly used KPIs to measure the success of automating the order to activate process.70%of CSPs that ranked the average handling time for B2C products/services as one of their top two KPIs for the order to activate process have seen a 2135%improvement in this KPI since 2020(Figure 4.5),an
115、d 60%expect an improvement of 3750%by 2024(Figure 4.6).Figure 4.5:Improvements in the KPIs that CSPs rank as one of their top two KPIs for the order to activate process since 2020,worldwide,202215 15 Question:“What KPIs are you using to measure the success of automation in the order to activate proc
116、ess?For your top two KPIs,what result have you observed since 2020?”;n=23.End-to-end automation:opportunities,challenges and the state of deployment|18 Analysys Mason Limited 2023 4:Measuring the success of automation Figure 4.6:Expected improvements in the KPIs that CSPs rank as one of their top tw
117、o KPIs for the order to activate process in the next 24 months,worldwide,202216 Problem to resolution The average time to complete a root-cause analysis is the most commonly used KPI to measure the success of automating the problem to resolution process,followed by the average response time(the amou
118、nt of time it takes to provide results to a request by user or customer),and then by both the average time to repair(also referred to as the mean time to resolution)and the average cost of equipment dispatched per service that did not resolve customer issues.CSPs that ranked any of these KPIs as one
119、 of their top two KPIs for the problem to resolution process have noted improvements of 150%since 2020(Figure 4.7),but improvements of at least 2135%and as high as 51%or more are expected in the next 24 months(Figure 4.8).16 Question:“What KPIs are you using to measure the success of automation in t
120、he order to activate process?For your top two KPIs,what is your desired goal for the next 24 months?”;n=23.End-to-end automation:opportunities,challenges and the state of deployment|19 Analysys Mason Limited 2023 4:Measuring the success of automation Figure 4.7:Improvements in the KPIs that CSPs ran
121、k as one of their top two KPIs for the problem to resolution process since 2020,worldwide,202217 Figure 4.8:Expected improvements in the KPIs that CSPs rank as one of their top two KPIs for the problem to resolution process in the next 24 months,worldwide,202218 17 Question:“What KPIs are you using
122、to measure the success of automation in the problem to resolution process?For your top two KPIs,what result have you observed since 2020?”;n=33.18 Question:“What KPIs are you using to measure the success of automation in the problem to resolution process?For your top two KPIs,what is your desired go
123、al for the next 24 months?”;n=33.End-to-end automation:opportunities,challenges and the state of deployment|20 Analysys Mason Limited 2023 4:Measuring the success of automation Termination to confirmation Most of the KPIs used to measure the success of automating the termination to confirmation proc
124、ess have shown significant improvements since 2020(Figure 4.9)and are expected to improve further in the next 24 months(Figure 4.10).Figure 4.9:Improvements in the KPIs that CSPs rank as one of their top two KPIs for the termination to confirmation process since 2020,worldwide,202219 19 Question:“Wh
125、at KPIs are you using to measure the success of automation in the termination to confirmation process?For your top two KPIs,what result have you observed since 2020?”;n=13.End-to-end automation:opportunities,challenges and the state of deployment|21 Analysys Mason Limited 2023 5:Recommendations Figu
126、re 4.10:Expected improvements in the KPIs that CSPs rank as one of their top two KPIs for the termination to confirmation process in the next 24 months,worldwide,202220 “We discussed with our partners,and the ones that are quite optimistic have a goal to achieve 80%automation.Today,our status is 50%
127、and our target is 70%,so it is feasible to reach 80-90%.Capex optimisation is the initial process,and then the opex.After an opex solution for the lifecycle,we have a cost estimate that we can save up to 2030%of our current cost on lifecycle management.”Director of IT/Network Strategy and Transforma
128、tion for the Middle East and Africa,Tier-1 CSP,Europe Expected improvement across all lifecycle processes It is worth highlighting that almost all CSPs expect to see further improvement in their automation KPIs in the next 24 months,regardless of the specific processes and KPIs.For example,90%of CSP
129、s that use the average TTM of B2B services as a KPI to measure the success of automating the service design to build process expect to see an improvement in this KPI of more than 20%in the next 24 months,compared to 80%of CSPs that have seen such improvements since 2020.5.Recommendations End-to-end
130、automation and orchestration of the network and service layers is a key priority for 5G operations,thereby allowing CSPs to accelerate the time to launch and flexibly scale up the monetisation and delivery of 20 Question:“What KPIs are you using to measure the success of automation in the terminatio
131、n to confirmation process?For your top two KPIs,what is your desired goal for the next 24 months?”;n=13.End-to-end automation:opportunities,challenges and the state of deployment|22 Analysys Mason Limited 2023 5:Recommendations new 5G services.Our recommendations for CSPs who are on their journey to
132、 full network automation are as follows.CSPs should increase their emphasis on automating the service design to build process using cloud-native CI/CD tools.Networks are becoming increasingly disaggregated and CSPs are faced with new challenges in streamlining their IT processes to support the testi
133、ng,acceptance and delivery of new network services,software and applications.CSPs should collaborate with vendors to implement an effective CI/CD pipeline by moving to agile NetDevOps approaches and automating manual processes for network updates,testing and validation.CSPs that wish to increase the
134、ir end-to-end business process automation capabilities should use components from a single vendor to integrate their assurance and orchestration processes.70%of respondents agreed that it is beneficial for assurance and orchestration processes to be integrated by the same vendor and most CSPs are cu
135、rrently deploying single-vendor VNF orchestrators.CSPs should select a technology partner that has the expertise to deliver all aspects of an end-to-end automation solution.The partner should provide business and design consulting,custom development and systems integration in order to manage and del
136、iver all aspects of the project,including solution integration and migration,requirements gathering and specification and process and workflow design.The partner should also support various deployment models for different modules or solutions,such as on-premises,cloud-native,managed service and SaaS
137、.Managed services enable the outsourcing of KPI management that many CSPs are trying to achieve.CSPs should deploy a vendor-agnostic,zero-touch automation solution to enable integration with third-party APIs and achieve a centralised network management and orchestration platform.Vendor-agnostic arch
138、itecture can deliver significant benefits to CSPs by simplifying operations and enabling programmatic network control and extreme automation across network domains and silos.This results in full NetDevOps transformation and aligns with the preference of 60%of the surveyed CSPs.This architectural app
139、roach provides more flexibility to select best-of-breed products and applications,and will in turn support CSPs objectives to reduce the TCO of existing networks.End-to-end automation:opportunities,challenges and the state of deployment|23 Analysys Mason Limited 2023 6:About the authors 6.About the
140、authors Justin van der Lande(Research Director)leads the Applications practice.He specialises in business intelligence and analytics tools,which are used in all telecoms business processes and systems.In addition,Justin provides technical expertise for Analysys Mason in consultancy and bespoke large
141、-scale custom research projects.He has more than 20 years experience in the communications industry in software development,marketing and research.He has held senior positions at NCR/AT&T,Micromuse(IBM),Granite Systems(Telcordia)and at the TM Forum.Justin holds a BSc in Management Science and Comput
142、er Studies from the University of Wales.Michelle Lam(Analyst)is a member of the Applications practice.She holds a BSc in physics and an MSc in quantum technologies from University College London(UCL),where she was an academic representative to the Students Union and assisted with research at the UCL
143、 Centre for Blockchain Technologies.She has also worked as a data analyst in machine learning and experimental physics,and has undertaken quantum computing research at the London Centre for Nanotechnology.Hansang(Andy)He(Manager)is a Manager in our Research division,based in London.He has advised va
144、rious clients across the telecoms value chain,but has a primary focus on equipment and software vendors including Amdocs,Ciena,Cisco,Ericsson,Fujitsu,Huawei,Nokia,Red Hat and VMware.Andy has recently managed engagements related to network automation and cloudification,artificial intelligence,private
145、 wireless networks and edge computing to support clients decision-making concerning product development,go-to-market strategies,and their establishment of thought leadership.Andy works closely with subject matter experts in our Research division to provide clients with recommendations that are based
146、 on market opportunity forecasts,competitive benchmarking and primary research entailing expert and executive interviews,as well as enterprise and end-user surveys.21 21 Analysys Mason Limited.Registered in England and Wales with company number 05177472.Registered office:North West Wing Bush House,A
147、ldwych,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 the results obtained from the use of this publication.The opinions expressed are those of the authors only.All information is provided“as is”,w
148、ith no guarantee of completeness or accuracy,and without warranty of any kind,express or implied,including,but not limited to warranties of performance,merchantability 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 i
149、n reliance on the information,including but not limited to investment decisions,or for any loss(including consequential,special or similar losses),even if advised 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.