Video analytics is a large and growing market

Video analytics is the processing and analysis of visual data (images or videos). When artificial intelligence is used to extract information from the data, it is referred to as intelligent video analytics or computer vision, although video analytics is often still used as a shorthand.

Video analytics stands out as a huge opportunity. It has the potential to be a killer application for edge computing, due to:

  • The large and growing market – In 2021 there were an estimated one billion surveillance cameras operational around the world. With the number of cameras predicted to grow by 20% in the period 2017-2024, AI and analytics will become increasingly important to capture value from the wealth of video footage being collected each day.
  • The ability for edge computing to grow the market – Without edge computing, video analytics is hindered by challenges with data sovereignty, and the cost of sending high-bandwidth data to the cloud (a problem that is heightened as video streams increase in quality). Edge computing therefore plays a key role in enabling video analytics, including more advanced AI/ML-enabled analytics, in a cost-effective way.
  • Its relevance to almost all industries – Video analytics can address a wide variety of use cases, from understanding consumer habits in retail, to analysing how football players kick a ball. In the case of video analytics for security, it is relevant across virtually all industries – education, transport, manufacturing, the list goes on.

The market for edge-enabled video analytics will be worth $75bn by 2030

Video analytics is a huge application for private 5G and edge computing, accounting for a quarter of edge revenues in 2021 (topped only by cloud gaming). In 2021 the edge-enabled video analytics market was worth over $5 billion globally. This is predicted to grow to $75 billion by 2030 at a CAGR of 34%.

video analytics

There are many application areas for video analytics, of which three are shown in the figure above. Of the three, video ingest and analysis for security and surveillance is the biggest short-term opportunity, representing an estimated 21% of the total edge computing market in 2021. This is due to the large base of installed security cameras that already exists, to which video analytics solutions can easily be retrofitted.

However, by 2030, video analytics for production and maintenance will be a larger opportunity. This will grow throughout the decade along with the move to Industry 4.0 and increase in automation resulting in an increase in sensors and analytics. Verticals like manufacturing, oil and gas and logistics will therefore be key adopters of this use case.

For more information about the video analytics opportunity at the edge, check our report How video analytics can kickstart the edge opportunity for telcos and STL Partners – Edge computing market sizing forecast


AI and the Future of Work

The Fourth Industrial Revolution is one of four major shifts that will have an impact on the Future of Work at telcos (others include societal and culture change, business environment change and pandemic related change)

The term Fourth Industrial Revolution is often used interchangeably with the technologies involved in Industry 4.0. However, in STL’s report The future of work: How AI can help telcos keep up, a broader definition is used (quoted from Salesforce):

“The blurring of boundaries between the physical, digital, and biological worlds. It’s a fusion of advances in artificial intelligence (AI), robotics, the Internet of Things (IoT), 3D printing, genetic engineering, quantum computing, and other technologies.” 

Analyst coverage of the Fourth Industrial Revolution in relation to the Future of Work focuses on the impact of new technologies on the economy, workforce and overall job market. These impacts are summarised below, and colour-coded based on the envisaged impact for telcos.

Expected impact of the Fourth Industrial Revolution

Source: Charlotte Patrick Consult, STL Partners

Economy

There are two scenarios for the world economy. The first is a boom/bust scenario caused by the rise of productivity, and in turn demand (the blue boxes), facilitated by the use of more technology and automation. The other is a period of lacklustre economic growth, which will follow if automation and technology adoption is slower. The second scenario is regarded as more likely – and the impact on telcos is expected to be less severe than the first.

Workforce

The skills shortage caused by the ongoing lag in government educational policy may be somewhat compensated for by machines in the mid-term. The Bain report Labor 2030: The Collision of Demographics, Automation and Inequality forecasts that a shortage of high-skilled workers will remain a significant issue for businesses.

Job market

The overall impact of new technology deployment may eliminate between 20% and 25% of current jobs according to MIT. But some job areas will experience growth.

Future of Work readiness

Telcos will have to respond to the changes introduced by the Fourth Industrial Revolution (and those introduced by the other shifts described above) to be ready for the “Future of Work”. Potential responses fall into three areas – strategic direction, skills development, and organisation and culture. Analytics, AI and automation (A3) tools can be useful in each. For example:

  • Data and analytics can help to improve organisational flexibility, particularly the speed of decision making in complex situations to inform strategic direction. The benefits of machine learning remain a promising future prospect.
  • More support from machines will be required to facilitate employee skills development (re-skill and upskill), plus onboard the increasing numbers of outside (contract) workers anticipated. Machines are also important to give workers the information they need to do their jobs.
  • Telcos will need to build trust levels around technology/A3 (algorithms to check machine decision making, explainable AI) to get humans and machines to work better together.

These are just a few of the ways in which A3 can help to tackle challenges of the Fourth Industrial Revolution and improve telco fitness for the Future of Work. For more, please see our report The future of work: How AI can help telcos keep up.

The Future of Work: How AI can help telcos keep up

What will the Future of Work look like?

The Future of Work is a complex mix of external and internal drivers which will exert pressure on the telco to change – both immediately and into the long-term. Drivers include government policy, general changes in cultural attitudes and new types of technology. For example, intelligent tools will see humans and machines working more closely together. AI and automation will be major drivers of change, but they are also tools to address the impact of this change.

AI and automation both drive and solve Future of Work challenges

Futuore of work AI automation analytics

Source: STL Partners

This report leverages secondary research from a variety of consultancies, research houses and academic institutions. It also builds on STL Partners’ previous research around the use of A3 and future new technologies in telecoms, as well as organisational learning to increase telco ability to absorb change and thrive in dynamic environments:

Enter your details below to request an extract of the report

The Future of Work

We begin by summarising secondary research around the Future of Work. Key topics we explore are:

Components of the Future of Work

Future of work equation

Source: STL Partners

  1. The term Fourth Industrial Revolution is often used interchangeably with the technologies involved in Industry 4.0. However, this report uses a broader definition (quoted from Salesforce):
    • “The blurring of boundaries between the physical, digital, and biological worlds. It’s a fusion of advances in artificial intelligence (AI), robotics, the Internet of Things (IoT), 3D printing, genetic engineering, quantum computing, and other technologies.” 
  2. Societal and cultural change includes changes in government and public attitude, particularly around climate change and issues of equality. It also includes changing attitudes of employees towards work.
  3. Business environment change encompasses a variety of topics around competitive dynamics (e.g. national versus global economies of scale) and changing market conditions, in particular with relation to changing corporate structures (hierarchies, team structures, employees versus contractors).
  4. Pandemic-related change: The move towards homeworking and hastening of some existing/new trends (e.g. automation, ecommerce).

Content

  • Executive Summary
  • Introduction
  • The Future of Work
    1. The Fourth Industrial Revolution
    2. Societal and cultural change
    3. Business environment change
    4. Pandemic-related change
  • How will FoW trends impact telcos in the next 5 to 10 years?
    • Expected market conditions
    • Implications for telcos’ strategic direction
    • Workforce and cultural change
  • Telco responses to FoW trends and how A3 can help
    • Strategic direction
    • Skills development
    • Organisational and cultural change
  • Appendix 1
  • Index

Related Research

 

Enter your details below to request an extract of the report

AI and automation priorities in customer care since the pandemic

This January 2022, STL Partners has updated its 2020 report A3 for telcos: Mapping the financial value, published in May 2020, which estimated the financial value of A3 (analytics, artificial intelligence and automation) through bottom up analysis of potential capex/opex savings or revenue uplift from integrating A3 into 150+ processes across a telco’s core areas such as:

  1. Networks and operations – BSS, OSS and networks;
  2. Customer channels – contact centre and digital channels, retail;
  3. Sales and marketing – up-sell, stimulation, CX and retention.

In terms of contact centre and digital channels…

A3 is used to tackle four types of problem:

  1. It improves the understanding of both customer and agent needs. For example, ML (machine learning) is used in segmentation and to trigger activities to allow next-best-offers (eg. a discount on insurance if the customer has just purchased a new handset from an agent). It also includes the use of sentiment and text analysis to monitoring agent quality.
  2. Understanding customer experience through the use of analytics and ML to collect and understand the large volumes of telco data.
  3. Understanding customer problems – both those immediately occurring and those expected to happen in future – through analytics and ML to understand the root cause, predict what might happen in future and prescribe either the best reactive or proactive action to be taken.
  4. All types of automation requirements

Our graphic summarises how the pandemic has affected key A3 priorities for telcos since 2020

Source: STL Partners

  • Case management includes use of RPA (Robotic process automation) and some ML to assign tickets and manage the process. A little more financial value has been assigned to the automation of the agent desktop as more case studies have emerged. However, in general A3 uses in this area are simple automation or analytics so account for a small proportion of savings.
  • Issue resolution is a related area and includes use of A3 to personalise offers or best-next-actions. These use cases are typically built around case management, workflow management and knowledge management and include a variety of personalisation capabilities – such as best next offer or action, problem diagnostics, proactive engagement and contextual routing.
  • Contact centre infrastructure has had a small uplift in its financial value due to new uses cases around call recording. Although, most of the value continues to come from the addition of A3 into IVR containment, where the financial benefit has been well understood over the last couple of years.
  • CX management: in recent research we highlighted that customer journey management will become more important as product sets increase in complexity with 5G.
    • Anecdotally, customer journey management tools haven’t always provided the expected utility for telcos, being difficult to use and lacking robust data feeds. However, vendors in this space are beginning to demonstrate new case studies.
    • A3 is increasingly used within the tools to improve data management, provide sentiment and text analysis to improve customer understanding and voice of the customer. They are also leveraging ML to improve customer journey management.

* The update to this research shows a market just beginning to mature and the financial model created in 2020 is holding up well. The underlying values from revenue uplift or capex/opex decreases have shifted up and down, but without radical change. The value is measured on an annual basis in dollar terms and as a proportion of total revenue for an “average telecoms operator”.

Coordination Age: Digital Twin applications for building construction

The Coordination Age is all about making better use of new and connected technologies to improve the management of resources. There are few industries that need it more than construction, which:

  • Generates 39% of industry CO2 emissions and where up to 30% of materials delivered to a construction site can end up as waste.
  • Employs 10% of the global workforce, where there are twice as many fatalities as most industries
  • Is worth $12Trn with an estimated inefficiency of 13% ($1.6Trn)
  • Has barely changed in productivity levels in 60 years between 1950 and 2012 (compared to say manufacturing, which has increased by 900% over the same period)

Construction Industry productivity 1950-2012

Digital Twin

Source: U.S. Bureau of Labor Statistics

As part of our ongoing research programme i spoke to Richard Ferris, Chief Technology and Product Officer at asBuilt, one of the largest independent building information modelling (BIM) consultancy specialists in Australasia. Richard was previously Group CTO at Lend Lease Group, a major ($10bn) Australian multinational construction, property and infrastructure company. He is also a leader in the development of digital twin technology (digital representations) in the architecture, engineering, construction and operation (AECO) value chain and founder and co-chair of the infrastructure working group for Digitial Twin Consortium.

What are the problems to be solved in construction?

Aside from the massive impact on carbon emissions from the production of concrete (which uses coal-based coking in its manufacturing process), there are numerous issues with the AECO value chain. For example:

  • designing complex, multi-dimensional sites in which the distribution of people, and use of energy, goods and services can vary considerably from moment to moment
  • managing the logistics and sustainability of resources
  • managing safety and operations on construction sites and in buildings
  • the energy efficiency and management of buildings and property estates

The construction industry itself is fragmented. There are a few large companies (like Richard’s former employer Lend Lease) and many thousands of contractors, sub-contractors, suppliers and roles where many work for different businesses or for themselves.

Each tends to be siloed into functional units, many are not highly advanced in their use of technologies, and the technologies themselves must be robust to survive the demands of a construction environment.

Information is generally highly siloed and may be stored in many different formats and file types. One area which has more uniformity is in billing, where companies have managed to digitise order and delivery forms to some extent.

However, the processing of bills and claims can still be both somewhat risky (things get lost / challenged, etc.) and generally stressful – after all, who doesn’t hate not getting paid?

What does asBuilt do to help?

asBuilt formally describes that its mission is to ‘harness 3D spatial intelligence to elevate people, performance and planet’. Richard calls it “a platform that is using spatial intelligence to improve the way the Construction industry manages information to support the generation and management of buildings and infrastructure”. 

To do this asBuilt has created a suite of propositions, such as Vault – a digital twin model of built assets, running on Microsoft Azure cloud. 

Despite his twin heritage, Richard prefers to use non-twin language explaining “it’s all about spatial intelligence. This allows people to see and find the information they need more easily. It allows customers to overlay different layers of information – for example environmental data superimposed on to reality captured from a building site and taking actions to manage that.”

asBuilt’s Vault proposition schematic

Digital Twin

Source: asBuilt

asBuilt solutions therefore aims to create a more intuitive and interactive data visualisation tool to help practitioners make better informed decisions.

Visualisation of asBuilt application on mobiles

Digital Twin

Source: asBuilt

A collection of case studies and videos are available here: https://www.asbuiltvault.com/construction. It’s notable that the New Zealand telecom operator Spark is involved in the first case study.

How might this develop in the Coordination Age?

asBuilt’s Vault proposition is a relatively new offering in the construction ecosystem and is the result of successfully using the platform to execute the company’s own building information modelling (BIM) services. 

A vast number of players are engaged in this activity, from classic IT/cloud and analytics giants Microsoft, Oracle, SAP etc., to IIoT players like Schneider Electric, and project collaboration and site progress specialists such as Procore. 

There are a vast range of approaches and architectures that can be applied and asBuilt appears to be an end-to-end solution designed to meet the needs of project management and construction companies. 

STL Partners has also seen de-centralised alternative design approaches by players such as Iotics who provide a versatile multi-party data sharing option. These approaches are not necessarily competitive but complimentary, allowing customers different ways to bring together different elements of the ecosystems they need.

Connected technologies can already help companies to go beyond ‘static’ historical data and connect sensors with ‘right time’ feeds so that decisions can be made in management situations as well as planning and design. 

It remains to be seen how impactful 5G will be this arena. Many challenges need to be resolved first, such as the willingness and ability of parties to share data and establishing trust in the security of connected devices and processes while simultaneously making devices discoverable to only those who should legitimately discover them.

To succeed, 5G connectivity (or 6G, Wi-Fi, Bluetooth, LORA, LPIoT, or whichever connectivity solution emerges in this arena) will need to be sufficiently versatile and reliable to connect in these environments, consume minimum energy, and operate at an infinitesimally small incremental cost.

Telcos may be able to play a role in facilitating the overall change. The obvious route is through their connectivity services where they can provide additional specialised and packaged services to the construction industry. To capture this opportunity they will first need to develop a more in-depth understanding of and the relationships within the AECO value chain.

What are digital twins

A digital twin technology is a digital representation of an existing physical or digital entity:

  • Examples of digital twins of physical entities include twins of simple sensors (such as a temperature sensor), machine components (such as a fan in a motor), a sub-system within a motor (such as a cooling system), the entire motor, or the whole vehicle containing the motor.
  • Examples of digital twins of digital entities include digital twins of data, a digital process (such as an order process or an automation protocol), or an entire digital business value network (such as a centralised data warehouse).

Digital twinning is a method of designing information systems that enables:

  • First visualisation, then dynamic control and emulation/simulation of assets. This can be ‘offline’ from the actual asset in the sense of a model to predict behaviours in different scenarios, or in real-time as a means to control and monitor operations.
  • A more efficient way to manage large volumes of data, where instead of collecting ‘data lakes’ storing every data point, data is organised into more manageable datasets capturing only meaningful events. This can reduce the need for data storage by up to 90%, which can be highly significant. An aircraft’s jet engine can generate Terabytes of data in a few hours of operation, for example. Customers often arrive at the need for digital twins with one or other of these needs in mind, and over time end up utilising both.

Where to use A3 for customer experience improvements

A3 (analytics, automation and AI) can be used to improve the customer experience and contribute financial value to telcos. Different types of A3 technologies are more or less important for enhancing different elements of customer experience. For example, analytics and machine learning (ML) can help to make sense of complex data, providing  insights into customer behaviour, preferences and experiences to increase customer understanding, while the use of bots and intelligence can remove routine work, speed up processes and increase quality. We identify the six main elements of customer experience below, highlighting areas where A3 can contribute meaningful value, across different functions.

A3 Customer experience

Source: STL Partners, Charlotte Patrick Consult

Four main themes for using A3 to to improve customer experience

A3 applications can be classified into four themes, as indicated in the diagram:

  1. Customer journey team: Telco teams that focus on individual customers could be equipped with suitable tools to understand and act on their issues. In the diagram, value-adding activity includes the addition of more machine learning to improve data management, the use of various AI techniques (such as sentiment and text analysis) to improve customer understanding and voice of the customer, and the use of ML to improve customer journey management tools.
  2. Automation: This is a broad category whereby automation can be used to speed up processes and transactions and improve accuracy, positively influencing customer experience.
  3. Personalisation: Another broad category which can be sub-divided further into two application areas, namely tools for marketing which allow more personalised recommendations, offers and actions (referred to as a “personalisation engine” in the diagram above) and tools for customer service, i.e. for personalisation of customer interactions in channels (the “customer engagement centre” above).
  4. AI: A collection of nascent tools which can solve specific customer experience issues and can provide better customer experiences in particular situations.

Telco progress on A3 for better customer experience

While telcos have made some progress in the application of A3 to improve customer experience, more could be done:

  • Analytics is commonplace for understanding the customer experience, but there is a delay in the application of ML. ML requires good quality data which can be difficult to obtain, particularly if it has to come from multiple channels. ML usage for customer experience across the first three themes is more limited than analytics usage.
  • In the early days of customer journey software (for understanding customer experience generally, rather than for understanding customer journeys across digital commerce), telcos often struggled to make good use of the insight provided, because it wasn’t understood by all teams which needed to use the product. Solutions to this issue include:
    • The requirement for product management roles and teams to assess particular journeys to ensure that there is an expert able to interpret the results and act on them
    • Continued work within these roles to increase the accuracy and relevancy of the customer journey maps created
    • A mix of technology and organisational change to allow access, use and sharing of journey maps
    • Improvement in data, processes and algorithms to create better insight. Telcos should especially focus on the use of ML in understanding patterns across very large data sets, where it will help to expose previously hidden issues.
  • So far telcos have only made limited progress in introducing additional data types from the network and OSS to enable views of experience with network, services, devices and applications to improve understanding of customer journey and personalisation of experiences in channel. There are a variety of vendors from the OSS space which have such products, but they are often a slow sell due to the need for the contact centre and other users to understand the benefits.
  • The box labelled “proactive technologies” on the diagram includes all solutions which use some form of personalisation to deliver proactive care or messaging to customers. For various reasons, it has been difficult to deliver certain types of troubleshooting for customers on-device. It is likely that the best solution to this will be to implement a mix of small proactive care solutions for particular customer issues and to focus on what can be easily delivered via new “assisted care” channels (such as messaging). For example, the pandemic has sparked the creation of services which connect technicians with broadband customers via video chat; the technicians can then use augmented reality (AR) to guide customers through device set-ups or resolve issues.

For more detail on how A3 can help telcos to improve customer experience, please see our report A3 in customer experience: Possibilities for personalisation

Related Research:

Advanced analytics for telecoms: how to unlock AI and ML capabilities

Advanced analytics for telecoms: how to unlock AI and ML capabilities

With more data than ever before, advanced analytics forms the base for AI and ML capabilities to play an essential role into the day-to-day operations of major telcos enabling smarter, more agile networks.

The value of advanced analytics for telcos

The explosion of sensors and the increasing comfort of executives to make data-informed decisions is driving new opportunities for telcos to utilize machine learning (ML) and artificial intelligence (AI). While 99% of operators surveyed in a recent STL report wanted to leverage these emerging technologies to increase network efficiency, different operators were at different points along their data analytics journey, with nearly all encountering obstacles.

However, in order to unlock the full potential of AI and ML in the long term, telecoms operators should focus on investing into developing their advanced analytics capabilities today, to form a strong foundation and base for future AI and ML skills and capabilities down the line.

Key opportunities for advanced analytics

STL Partners identified six types of problems that A3 (automation, advanced analytics and AI) can help with and conducted a modelling exercise to assess the value of A3 into a telco’s processes. Our analysis predicted that 60% of the value from A3 will come from the network domain.

One key problem that telcos face is making sense of complex data, where using advanced analytics and machine learning can help to identify patterns, diagnose problems and predict/prescript resolutions.
Within this domain, there are two key use cases where we report the largest potential financial benefit (i.e. more than $50mn in yearly financial benefit):

1. Network (resource) management
2. Service assurance

Network management

Among the most important opportunities identified is network (resource) management, which includes network planning, deployment and maintenance, as well as the management of network capacity and resources. This is no particular surprise given that the management of physical and virtual network infrastructure accounts for a significant proportion of operators’ capital and operational expenditure. A3, including advanced analytics in particular, can play an important role in enabling operators to leverage network insights and data to make better informed decisions around new network investments. Predictive maintenance has also singled out by operators as an important use case that will enable them to more efficiently allocate resources to repair their networks, substantially lowering OpEx.

Service assurance

This is another key area where advanced analytics (and eventually ML and AI) can play a significant role, particularly in the short term. This is also heavily intertwined with the virtualisation and cloudification of networks (and network functions) whereby new service assurance products will be required in the next 5-10 years to support an increasingly multi-vendor, disaggregated environment as well as new types of edge computing services and maturing IoT use cases. That is not to say that the value of advanced analytics is primarily in the “new” stuff, our research has found that there is still value that has not been realised and can be derived from A3 in service assurance for 3G and 4G LTE, for example with the introduction of more predictive algorithms. Advanced analytics in service assurance will set up greater possibilities with the use of machine learning further down the line for more proactive root cause analysis.

Next steps: preparing for greater AI and ML opportunities

Ensuring that data is clean and unified

The majority of operators cited that data collection and management is still a key issue and challenge, therefore operators should focus on improving this to ensure that their data is clean, complete and unified, and that their data lakes are up-to-date and accurate.
Building analytics skills and capabilities

Telecoms operators should look to build the skills internally to make full use of their data, whether that be through hiring more data scientists or upskilling existing employees. Data analytics should not be seen purely as a capability to leverage internally, for example within the network, but also seen as one that others can leverage (i.e. moving to charging customers for outcomes and insights from analytics).

Creating a data-centric culture

To create a data-centric culture telcos should consider creating a Chief Data Officer that takes ownership of data collection, aggregation, cleaning, and data analysis. This helps to create a data-centric culture by having an advocate for analysts on the board. Talent acquisition and retention of individuals with key analytics skills should also be prioritized.

ML/AI provide telcos with exciting new opportunities, but it will take time, especially for lower-tier operators, to see them bear fruit.

Author: Ciarán Mulqueen is a Consultant at STL Partners, specialising in telco-cloud

Read more about Telco Cloud, Cloud Native, NFV & SDN

Research

NFV goes mainstream: How cloud-native is contributing to growth

The number of deployments of NFV and SDN continues to grow – but while some markets are moving on to the next phase, others are just getting going.

Read more

Research

Open RAN: What should telcos do?

Alongside the roll-out of 5G cores and radios, the Radio Access Network (RAN) is evolving to a more open, virtualised and distributed architecture. What are the opportunities and risks for telcos?

Register now

Research

Telco Cloud: Why it hasn’t delivered, and what must change for 5G

Telco cloud made big promises for the transformation of telecoms. It is a fundamental enabler for 5G and the exciting opportunities ahead. Why hasn’t it delivered yet – and what needs to change?

Read more

Achieving eCommerce best practice through data advancements

There are clear benefits to operators who successfully implement A3 (analytics, artificial intelligence and automation) within their organisation.

  • Analytics: Collecting and analysing data to drive better decision-making
  • Artificial intelligence: Using computing capabilities to perform human cognitive functions
  • Automation: Replacing or supporting activities that require human input with machines

This is evident in the industry’s investment in multiple telco domains, albeit with varying levels of maturity. Example use cases include:

  • Personalised marketing and call centre augmentation in sales and marketing
  • Customer value management and troubleshooting in customer experience
  • Infrastructre planning and predictive maintenance in network planning and operations
  • AI-as-a-service and product development and augmentation in service innovation
  • Supply chain management and SLA risk modelling in other operations

For more information, check our report A3 for telcos: Mapping the financial value

 

AI for network-driven customer experience

Although customer-facing teams in marketing and sales have been quicker to adopt AI and automation than network operations teams, there is a growing understanding in telecoms that delivering a good customer experience must start with the network. Achieving this is easier said than done, since getting a full view of an individual customer’s experience of the network depends on bringing together data from potentially hundreds of systems, and then making sense of it. Figuring which metrics are most representative of a customer’s experience and how to combine and present them in an easy-to-understand way for network operations and customer-facing teams is one of the biggest challenges operators are grappling with today. However, it is well worth it – our research into the value of A3 (AI, analytics and automation) for telcos shows that improvement to top level customer experience from using A3 for network service assurance is worth US$27 million annually for the average sized telco.

How O2 is using AI to improve customer experience

O2's network customer experience metric

O2 introduced a single data driven customer experience metric called NCX (network customer experience). By using machine learning techniques to evaluate a customer’s perception of network quality, O2 were able to impact three key areas:

  1. Network operations: NCX supports O2 in network fault management and network optimisation to better invest its resources into improved fault detection, faster root-cause identification, smarter dispatch of engineers into the field.
  2. Network planning and investment efficiency: O2, like its peers, is under pressure to constrain its capex. NCX allows the organisation to take a nuanced view of CX across different parts of the network and determine which should receive investment for largest CX improvement ROI.
  3. Targeted marketing and CX management: The granular information NCX provides about individual customers allows O2 take a more tailored approach to communication. Shift from reactive to proactive has helped create a ‘stickier’ network, increase customer retention and improve upsell opportunities.

Ultimately, implementing a network-driven approach to customer experience management can feed into telcos’ customer-facing channels, enabling them to implement a successful omnichannel strategy.

See our in-depth research on A3 in customer experience

 

Automation in the telecoms industry: A key differentiator

Automation in the telecoms industry: A key differentiator

AI and automation are becoming increasingly important to telco strategies as they seek to handle new network complexities and address revenue decline. This article explores how telcos are thinking about automation, and why they should think about moving from intra-process automation in the short term, to inter-process automation in the longer term.

Automation will be a key pillar for telcos to achieve competitive differentiation in an increasingly saturated market. Whether their focus is on providing exceptional connectivity, or they are seeking to offer new solutions beyond connectivity, telcos will need to introduce more automation into their network and service operations to address shrinking revenues from network commoditisation.

STL Partners recently surveyed more than 100 key individuals from telecoms operators globally to understand how telcos are pursuing automation and defining their strategies. Based on this, and interviews with 15 leading operators, we produced a set of recommendations for telcos in defining their automation strategies. Our report ‘Prioritising automation: Creating a successful building block strategy’ summarises our key findings and looks at the following aspects:

  • How automation, AI and data analytics are driving telco strategies
  • How telcos can create a building block strategy to succeed with automation
  • How telcos can begin to innovate with automation, from people & culture to systems & technology

Why do telcos need automation?

70% of our survey respondents stated that they intended to move beyond providing basic connectivity and were seeking new opportunities for revenue growth. Increasingly, telcos are seeking to exploit emerging technologies such as 5G and edge computing to offer differentiated solutions to customers, particularly in the B2B2X space. However, network and service automation will be a key enabler for all telcos to address declining revenue growth, regardless of their strategic approach.

Figure 1: Automation will be critical for telcos whether they are focusing on connectivity, or exploring new avenues for revenue growth

Automation in telecom industry

AI/ML-based automation is much more nascent (mostly at POC stage) for telcos than rules-based automation which has long been part of network processes. However, as outlined in our previous research, STL Partners views telco automation journeys as a continuum, wherein telcos should gradually transition from rules-based (business intelligence and fixed-policy automation), to ML-supported automation and ultimately fully autonomous systems that include self-improving and self-learning capabilities.

In order to provide strong connectivity, manage increasing network complexities (with 5G), and maintain cost efficiencies, telcos will need to define a roadmap for how they can introduce more automation into their network and service operations. For telcos seeking to exploit new technologies, more intelligent automated processes (that leverage AI/ML) will be a critical enabler – they will struggle to rely on manual processes to manage increasing complexities:

  • Network operations: on the network side, use cases such as network maintenance, fault detection, self-optimisation and CI/CD will be essential to manage increasing complexities – for example, telcos might need to manage network functions across multiple locations (e.g. with edge computing)
  • Service operations: telcos seeking to move higher up the value chain will need to explore dynamic business models that bring in additional complexities – use cases such as customer and partner management, service provisioning and service assurance will be key enablers for this.

Figure 2: Automation of network and service operations could bring big financial benefits for telcos, especially compared to automation of other functions

Automation in telecom industry

The current state of automation in the industry

However today, most telcos are far away from deploying this level of “full automation” and are implementing specific use cases according to their unique needs (e.g. maturity, geographical or network focus). In our survey, <10% of telcos report being fully automated in any single domain within network and service operations. Automation efforts have been much more focused on non-network domains such as HR, sales, or marketing. Within network and service automation, most adoption is in use cases where vendor solutions are mature and widely available (e.g. customer billing and revenue management) compared with those that are more nascent (e.g. network function & lifecycle management).

Though all telcos have some level of automation, this is mostly “intra-process” (automation of unique functions) – based on our survey, telcos on average have automated ~40% of processes within single domain use cases. To successfully compete and exploit new technologies in the long term, telcos will need to move towards “inter-process” automation (ultimately closed-loop automated processes spanning multiple domains). For example, as telcos increasingly adopt cloud-native models with microservices, it will be incredibly difficult and inefficient for all network activity to operate and be controlled separately. Network functions will require more autonomy and need to be able to communicate with each other, particularly as data from different parts of the network is brought together for more intelligent decision making (e.g. orchestration and assurance).

Figure 3: Intra-process automation will be a starting point, but telcos should begin to define their roadmap for inter-process automation

Automation in telecom industry

As outlined in our report, A3 for telcos: Mapping the financial value, telcos stand to save ~5.7% of their total annual revenues by automating their network and service operations. As proof points emerge and technologies mature, we expect to see more adoption of intelligent automation within the telecoms industry. Intra-process automation is an obvious starting point and an area that all telcos should be pursuing to build up the technical capabilities and expertise to support individual automation use cases, but inter-process automation should be viewed as the ultimate goal to succeed in the long-term.

Author: Reah Jamnadass is a Consultant at STL Partners

Telco Cloud insights pack

This 24-page document will provide you with a summary of our insights from our virtualisation research and consulting work:

  • Overview of Telco Cloud deployments worldwide
  • Benefits of telco cloud: state of the industry
  • Deployment approaches: implications and challenges
  • How STL Partners can support you

Just click on the button below to request your free pack.

 

Read more about Telco Cloud, Cloud Native, NFV & SDN

Research

NFV goes mainstream: How cloud-native is contributing to growth

The number of deployments of NFV and SDN continues to grow – but while some markets are moving on to the next phase, others are just getting going.

Read more

Research

Open RAN: What should telcos do?

Alongside the roll-out of 5G cores and radios, the Radio Access Network (RAN) is evolving to a more open, virtualised and distributed architecture. What are the opportunities and risks for telcos?

Register now

Research

Telco Cloud: Why it hasn’t delivered, and what must change for 5G

Telco cloud made big promises for the transformation of telecoms. It is a fundamental enabler for 5G and the exciting opportunities ahead. Why hasn’t it delivered yet – and what needs to change?

Read more

data monetisation in telecoms: 10 use cases

value of telco data products by verticals

Big data monetisation in telecoms has been an area of activity for the last few years. However, telcos’ interest levels have varied over time due to the complexity of delivering and selling such a diverse range of products, as well as highly variable revenue opportunities depending on the vertical. Telcos’ appetite to pursue data monetisation strategies has also been heavily impacted by the fortunes of other new telco products, in particular IoT, owing to the link between many telco data and analytics products and IoT solutions. As illustrated in the graphic above, in many sectors IoT data monetisation is the main strategy, while in others telecoms operators can address opportunitinities independently of IoT services.

We explore the main data monetisation models and use cases across 10 verticals. We break these up into ones where data monetisation strategies are strongly linked to IoT, and those that are more independent.

IoT data monetisation opportunities

1. Agriculture

Most activity in the agriculture sector is seen from large multi-national telcos with mature IoT propositions. Not all the biggest telcos report pursuing such projects, though, with case studies most likely from those with a strong presence in developing markets, or with large multinational enterprise customers present in developing markets.

Most opportunities are related to IoT and sensors and include a mix of connectivity services with storage and analytics of the payload data. For more complex and specialist the use cases, telcos are much more likely to play a connectivity only role. For example, in crop management, NTT Docomo offers hardware and analytics, but many other telcos instead choose to work with specialist platform vendors.

2. Manufacturing

Much of the discussion about future telco activity in this vertical is linked to the provision of 5G services to allow Industry 4.0 capabilities1 . The most visible telco manufacturing solutions are often linked to historic associations with a particular industry; for example, Vodafone and T-Systems solutions within the automotive industry. Barriers for telcos to overcome include rolling out 5G capabilities fast enough to satisfy manufacturers and enable the swap out of LTE, creation of flexibility in their offerings and ease of access through on demand provisioning etc.

The table suggests that there is very little financial value for data/analytics in the vertical, but this is linked to the prevalence of IoT use cases where data analytics will not be sold as a separate service. It is likely that telcos which choose to focus aggressively on 5G and edge computing for manufacturing are most likely to take advantage of the data/analytics opportunities – predictive maintenance and the provision of analytics for autonomous vehicles on the factory floor look most promising.

Some of the solutions where telcos are most active in manufacturing, such as asset management, supply chain analytics and transportation/logistics solutions, are also provided to other verticals. These are therefore captured in the section considering horizontal solutions for all verticals.

3. Retail

Historically, this was one of the first verticals targeted by telcos with customer movement inside products. Developing the products was often hampered by the difficulties of finding the right person in a retail organisation and the likelihood of non-standard requirements from every retail customer. However, among larger telcos with ambitions in data/analytics there is now a reasonably mature retail product set.

Ongoing opportunities divide into three categories:

  • Customer movement insight products: These tend to be the most feasible project as they are more mature and use telco data, for example for store placement calculations.
  • Customer insight products: Related projects use customer insight (demographic, sociographic) rather than geolocation data. For example, the open data platform described above could be accessed by retailers, hoteliers or other types of customer in this vertical.
  • IoT/small cell opportunities: There are additional data/analytics opportunities which use small cell, video and CCTV data to track customers in small spaces or within a shopping mall – however, these are considered of lower feasibility because they require rollouts of these capabilities and potentially IoT related products such as sensors. These opportunities subdivide between those that require specialist analytics and those that require additional AI capabilities such as facial recognition. All of these use cases require a sustained focus on the retail sector and its needs, plus enough rollouts of small cells, wifi, beacons etc to make a business case for adding data/analytics on top.

4. Transportation

Like other verticals, most of the most accessible financial opportunity is from customer movement insight provided to passenger transport companies such as trains and buses. This is a reasonably mature use case for telcos. Much of the rest of the opportunity is related to mature fleet management markets where there are limited opportunities for adding data/analytics. Lastly the connected vehicle market provides various potentially feasible opportunities to add data/analytics to IoT deployments.

Independent data monetisation in telecoms

5. Finance

The feasibility of providing services for retail and investment banks and other companies within the financial services sector divides broadly into three categories:

  • Services live today: anecdotally, location-based card authentication (i.e. alerting a bank when a customer travels to a different country, which improves fraud management) is one of the highest revenue services for telcos today. There are additional services alerting retail banks to potentially fraudulent behaviours, but these seem less popular. Services using customer movement insight such as identification of where to open a bank branch are also popular, although the financial benefit is not seen in the table below as it is categorised with other similar services for other high street retailers.
  • Possible services not yet on the market: customer movement insight could also be used for optimising the location of bank ATMs and telco data could be added to specialist analytics for operating them, however example services have not yet been seen from telcos, so it is possible that there is limited demand.
  • Specialist services: data and analytics services on high speed, complex customer and market data which offer less attractive opportunities for telco services, but is not completely infeasible. For example, there are cases of telcos adding customer movement insight data to improve bank trading decisions and risk management. There are also examples of telcos, such as CenturyLink, who have purchased analytics companies because they host financial data, although it is not clear how much financial return this has delivered for them.

6. Insurance

Insurers use external data for risk management, actuarial calculation and underwriting decisions. There are compelling reasons for insurers to include new data sources, however, there are regulatory restrictions (as companies need data on individuals) and it needs to be verifiable and up to date. There has been very little telco activity in this field – except for the odd anecdotal data point that they may be working with specialist actuarial consultancies. The financial value ascribed to the provision of data is therefore mostly for niche products that do not need PII, while likelihood scores are low as the limited opportunity means a reduced sales focus for operators.

One area of insurance where there has been strong telco involvement is in telematics products for insurers, including usage-based insurance. Analytics create driver scores for pricing and risk management purposes. Tier 1 telcos including Verizon, Telefónica, Telstra and Orange have data monetisation products in this area – some create the analytics themselves while others partner.

7. Healthcare

Building new revenues in the healthcare vertical requires telcos to have a long-term strategy and a real understanding of the sector. From a data and analytics perspective, nearly all telco activities include the transport and storage of data. However, they also then require a mix of specific platforms, applications and smart devices dependent on the use case, which potentially offer the opportunity for addition of A3 (automation, analytics and AI). As the market matures, different strategies are seen towards investment (build or buy) up the value chain which allow telcos to develop A3 capabilities.

The opportunities divide into various categories:

  • Telemedicine use cases provide smart devices which generate payload data. The data requires transportation and storage, also providing opportunities for development of analytics to generate alerts or provide historical trends.
  • The management of electronic health records, medical images, electronic prescriptions and insurance claims. These require data transport, storage and then specific platforms for exchanging information between different parties.
  • Solutions for the pharmaceutical and life science industries including collaboration platforms for clinical trials.

 

8. Real estate and construction

This vertical offers a number of opportunities for customer movement insight products. Anecdotally, deal sizes are smaller than in, say, retail, although location mapping is useful for a variety of purposes. Use cases require a good deal of external data and open data from government platforms to be successful.

Potential opportunities include:

  • Use of customer movement insight to understand demographics, behaviours and requirements of a local community to improve development and investment decisions for both retail and commercial real estate companies
  • Use of the data for pricing, marketing and sales decisions within estate agents and brokers
  • Use of indoor data from small cell deployments within shopping malls to understand customer movement in order to position advertising, adapt opening hours according to foot traffic and change layouts to drive traffic to, say, food courts.

9. Telecom, media and technology

Provision of insight to entertainment/sporting venues is a relatively common use case today that uses customer movement insight and sensor data. There is also opportunity for analytics such as customer segmentation and behaviour. Projects telcos have reported participating in tend to include a significant consulting component, so this is best suited to operators with a consultancy team.

Other opportunities around content consumption patterns are more difficult for telcos. Telcos may well have insight from their set-top boxes and other platforms that will be of interest to content providers, but it is a mature market which is used to ingesting different types of data and it does not seem a popular use case.

10. Utilities

This market is split between products for consumers which seem to be increasingly hard for telcos to deliver, A review of telco websites suggests that, except for a couple of exceptions, most have retreated away from a variety of smart home products towards a focus on security. (STL has previously argued that the smart home in itself is not a viable product, but rather that telcos should focus on solving specific issues for households, such as security, entertainment, or energy efficiency. See STL report Can telcos create a compelling smart home?)

Products for the utilities themselves are mature and larger telcos have been successful. Telcos offer a range of monitoring and management capabilities for the grid and smart meters, with additional products including security, communication networking solutions, drone management and fleet management. There are three main categories of products in which customer movement insight data could be included alongside analytical solution using IoT payload data:

  • Grid distribution, monitoring and control: the largest telcos offer descriptive and diagnostic analytics on data about electricity, water and gas networks. 5G will offer new opportunities for real-time prescriptive activity using digital twins. Meanwhile, shifting the energy market from fossil fuels to renewables will require matching demand to supply (when the sun shines and the wind blows), as opposed to the current environment of matching supply to demand whenever it occurs, which will in turn require very advanced analytics and automation across all levels of the energy market.
  • Smart metering control and management: currently a mature market, with opportunities to add prescriptive analytics that enable better management of problems. This area will also evolve significantly over the coming decades towards smart “just in time” energy usage in homes and businesses.
  • Site and network planning: Opportunities for customer movement insight data to be added to give information about the population to enable new installations (pylons, sub-stations, water facilities, green-energy installations etc).

For more information, check our report Telco data monetisation: What’s it worth?

 

A3 technology: Where should telcos focus?

Prioritisation matrix for telco A3 capabilitiesA3-AI-Automation-Analytics

STL recently explored potential enterprise solutions leveraging analytics, AI and automation (A3) capabilities that telcos can address across 200+ use cases across 14 industry verticals.

We looked at vertical opportunities and sought to answer two questions: Which of the A3 technologies are most feasible for telcos to work with? and which verticals hold the most opportunity?

In terms of A3 technologies available to telcos today, we asked: How feasible is it for a telco to deploy the A3 technology successfully? (y-axis) and how many opportunities are there for the particular technology to be deployed across all the vertical use cases? (x-axis).  See graphic above.

The key takeaways from this element of our analysis found:

  • Pattern and anomaly detection and personalisation are by far the most numerous and attractive opportunities for telcos. The addition of analytics and machine learning to track behaviours in both people and things often provides input into forecasting and optimisation exercises. They also find patterns or triggers in order to suggest actions to be taken.
  • Image and speech recognition, also offer a good number of reasonably feasible opportunities, particularly for large telcos with related products and good traction in the relevant verticals (retail, real estate and construction, manufacturing).
  • “Augmented workers” capabilities is a more speculative addition to the graphic. These technologies work alongside humans to provide decision support where large data sets and complex decisions are needed. We include these capabilities as they could be added to telco analytics solutions to provide additional functionality to customers in future, particularly in contact centres.
  • Text, sentiment and emotion detection are mostly unattractive opportunities for telcos, tending to be used in use cases that are of low interest to them.
  • Immersive technologies and AI design are considered of lower feasibility currently due to the immaturity of the technologies and markets.

This report considers in depth each of these different types of A3 capabilities and their uses for a telco as well as where it is most feasible for telcos to provide A3 technologies to different verticals.

Download the report here: A3 for enterprise: Where should telcos focus?

Related research report:

The report builds on a previous STL Partners report Telco data monetisation: What’s it worth? which modelled the financial opportunity for telco data monetisation – i.e. purely the machine learning (ML) and analytics component of A3 – for 200+ use cases across 13 verticals.

Telecoms network automation: APAC prioritises innovation

Telecoms network automation

Source: STL Partners telecoms survey, March 2021, 100 respondents

According to our survey, automation is primarily driven by the need to reduce the operating costs of existing networks and services for operators in EMEA and North America. In APAC, on the other hand, automation adoption tends to be a necessary consequence of network development as they develop new network services that cannot run without automation.

Many operators reported prioritising automation domains that can help them optimise internal efficiency so that they can focus on developing new services and revenue streams. One APAC operator suggested that automation of network and service domains could free their operations team from having to handle immediate problems with the network, leaving them more time to engage with customers and focus on service innovation.

Differences in automation maturity, innovation budgets, and C-suite ambitions will all impact how operators prioritise domains. Operators vary greatly in how they prioritise “explore or exploit” options – namely, building on existing strengths versus investing in higher-risk and potentially higher-reward development. In the short term, operators with legacy networks may need to prioritise streamlining those with automation over developing emerging use cases.

Telecoms operators’ network automation aspirations are also shaped by market positioning; new entrants will not have to undergo as significant a journey as incumbents who need to reshape their employees, technical capabilities and operating model. For new-entrant operators, automation is often at the core of their business strategy, enabling them to quickly build out, scale, and manage a cloud-native network with a limited workforce. One APAC challenger operator described automation and a DevOps approach as being essential to their ability to roll out a greenfield network and accelerate 4G site commissioning from 3-4 days to 8 minutes.

See our in-depth research on AI & automation in telecoms:

Why AI in telecoms matters in the Coordination Age

Why AI in telecoms matters in the Coordination Age

As we move into the Coordination Age, telcos are under increasing pressure to innovate and find new sources of revenue growth. AI is one of a host of new technologies which can help them to achieve this.

At STL Partners we speak of the Coordination Age, the third age of telecoms. The first age of telecoms was the Communication Age and was about connecting people, first through telegraph and then through telephony. The Information Age was the second age of telecoms and began with the inception of the Internet. The Coordination Age will be defined by an increase in the number of devices that are connected as IoT becomes widespread, this will lead to a massive increase in data volumes and a need for greater resource efficiency.

Figure 1: The Coordination Age

AI in telecoms

Source: STL Partners

The Coordination Age coincides with a time of stagnating telco revenue growth. Telcos need to find new channels of revenue growth and move to a more decentralised B2B2X business model. 5G will be a key technology for helping telcos to re-invigorate revenue growth, but there will also be a key role for other technologies. Artificial intelligence should be at the centre of telco strategy in the Coordination Age.

Artificial intelligence (AI) is the ability of a computer or a robot controlled by a computer to do tasks that would usually be carried out by humans as they need human intelligence or decision making.

AI can be applied at all stages of a telco’s operations. It will allow telcos to better utilise the vast wealth of data that is available to them. As other technology is adopted (5G standalone, Open RAN, IoT, and the move to edge computing and greater automation) the role of AI will only grow, enabling telcos to manage and optimise operations, as well as to plan for future deployments more easily. It should enable greater automation which is integral to the Coordination Age.

There are many use cases for AI in telecoms and in this article we run through some of those that we believe will be key for telcos in the Coordination Age.

AI use cases for telco networks

STL Partners has previously written several reports looking at AI in telecoms, including this report looking at AI use cases in the network. We split the use cases into 3 main categories:

  1. Fault detection, prediction, and resolution: speeding up the process of identifying and resolving network faults, including predictive maintenance. Having a functioning network is vital when seeking to move to a B2B2X model.
  2. Network optimisation: routing traffic and balancing workloads across telco infrastructure to deliver a cost-effective service at the highest quality possible. While this could be a manual process, automation allows network engineers to allocate time to other important areas.
  3. Network planning and upgrades: this will be particularly important with the rollout of 5G. Making sure that networks are deployed in an optimised and efficient way is a key aspect of the Coordination Age.

There are a number of sub-use cases within each of these 3 overarching categories.

Within fault detection, prediction, and resolution, root cause analysis is an important use case. There is a huge bank of data on historical network faults and machine learning (ML) models and AI can learn to predict future network faults. Even for newer sources of network faults, AI can work to ensure processes predict faults beyond what current models or network engineers could. AI can also help with fixing the actual problems. Using this huge bank of data AI models can calculate different solutions by speed, cost, or impact on customer experience. Again, this will save network engineers time and resources.

Use cases within network optimisation are likely to target either cost or quality of service. Regarding cost, AI may be used to maximise the use of existing network assets or perhaps to prioritise and plan firmware updates for times that will be least disruptive for customers. For quality of service, AI may optimise the distribution of traffic between cell sites, maybe due to service disruptions, and then revert to the default configuration after the event or peak is over.

Within network planning there are a multitude of potential use cases for AI. For example, AI can map rural populations with real network coverage to identify underserved areas, Telefónica have already deployed it in this way (see figure below). This may be used for 4G networks in less developed countries, but it has its uses for 5G as well. Factories or other commercial sites outside of urban areas could be mapped to ensure that 5G coverage includes these areas that are likely to take advantage of IoT and other applications. It is important for telcos to enable applications such as these as we move into the Coordination Age.

Figure 2: Telefónica have used AI for network planning and maintenance

AI in telecoms

Source: STL Partners

Other AI use cases

There are use cases that go beyond telco networks. We now run through a couple of the most prominent examples.

AI in customer service

Virtual Assist for Customer Support is one of the most popular uses of AI across industries, not just for telecoms. There has never been a greater demand for telecommunication services, not just due to the Covid-19 pandemic but also by virtue of the fact that more and more of the world’s population is gaining access to technology and connectivity. This will only increase as 5G causes more and more devices to be connected, hence the Coordination Age is defined by increasing data volumes.

This increase in data volumes will inevitably lead to an increase in demand for customer service within telecommunications. As highlighted earlier in this article, AI will be able to minimise the faults in the network, but it can help in other ways. By equipping customer support lines, whether online chats or over the phone, with AI you will not only improve customer satisfaction by cutting wait times down, but telcos can also save costs by using fewer human operators. AI could either support a chatbot and hence resolve queries automatedly, or it can support customer service employees with information such as next best action to solve a customer’s query or routing customers to the most appropriate agent. There is increasing appetite for virtual assistants, indeed 74% of organisations view conversational intelligent virtual assistants as an important enabler of successful customer engagement.

AI in marketing

AI is also increasingly being put to use in marketing. Telcos should seek to take advantage of these use cases as well, Google research found that 90% of marketers believed that personalisation significantly contributes to business profitability. The massive amount of data that telcos have available on customers puts them in a good position to personalise their marketing for individual customers. The use cases may include:

  • Personalised ads and messaging: if you can automate the process of targeting the right customers with the right content when they need it, this should bring increased sales.
  • Customer segmentation: AI can help telcos to obtain more granular segmentation for groups of customers.
  • Predictive analytics: this should help with retaining customers or pushing them onto more profitable solutions for the telco.

Data monetisation

Another area where there are a lot of use cases is in telco data monetisation. This involves telcos selling their data or insights from their data to third parties. These third parties will often be enterprise customers, and previous STL Partners research looked at 200+ use cases spread across twelve different verticals within this space.

Robotic Process Automation

An automation use case that can run along AI use cases is Robotic Process Automation (RPA). This involves the automation of rules-based business processes, configuring software to capture and interpret applications for processing a transaction, manipulating data, triggering responses, and communicating with other digital systems. Essentially it means getting software to emulate the actions of a human interacting with digital systems to carry out business processes. This should enable telcos to reduce costs and increase efficiency.

Why do telcos need AI?

These use cases and many more that have not been mentioned in this article will be essential for telcos in the Coordination Age. There is a need for innovation to encourage revenue growth and ensure that enterprise customers do not move to the hyperscalers and other big tech players. Enterprises need automated, programmable networks that are highly flexible and adaptable to a wide range of customer requirements. If telcos do not sustain momentum in implementing AI and automation in their networks and services, then others including tech players will find ways around that, as they did in 4G by running services independently from networks.

These groups of competition are already using AI on a wide scale, but even if telcos do not feel under pressure from external groups, they should be wary of within their own market. Telcos who buy into new technologies including AI will gain an advantage and may pull ahead of competitors. Indeed, according to this report by Anodot, CSPs that are serious AI adopters with proactive strategies report current profit margins that are 5 – 7 percentage points higher than the industry average.

AI will be one of a number of new technologies that telcos will need to use as we enter the Coordination Age. In a previous report, STL Partners estimates that telcos can save up to 7% of annual revenues through the adoption of AI, automation, and analytics. Telcos cannot afford to miss out on these benefits.

Author: Matt Bamforth is a consultant at STL Partners, specialising in telco cloud, edge computing and 5G

Telco Cloud insights pack

This 24-page document will provide you with a summary of our insights from our virtualisation research and consulting work:

  • Overview of Telco Cloud deployments worldwide
  • Benefits of telco cloud: state of the industry
  • Deployment approaches: implications and challenges
  • How STL Partners can support you

Just click on the button below to request your free pack.

 

Read more about Telco Cloud, Cloud Native, NFV & SDN

Research

NFV goes mainstream: How cloud-native is contributing to growth

The number of deployments of NFV and SDN continues to grow – but while some markets are moving on to the next phase, others are just getting going.

Read more

Research

Open RAN: What should telcos do?

Alongside the roll-out of 5G cores and radios, the Radio Access Network (RAN) is evolving to a more open, virtualised and distributed architecture. What are the opportunities and risks for telcos?

Register now

Research

Telco Cloud: Why it hasn’t delivered, and what must change for 5G

Telco cloud made big promises for the transformation of telecoms. It is a fundamental enabler for 5G and the exciting opportunities ahead. Why hasn’t it delivered yet – and what needs to change?

Read more

Skilling up telcos for A3: New roles and where they fit (Chart)

Our recent research explored the steps telcos are looking to take and should take in order to pursue data-driven strategies in automation, analytics and AI (A3).  Our surveys on industry priorities suggest that operators recognise this need, regardless of whether they are focused on their core connectivity business or seeking to build new value beyond connectivity. A corresponding organisational shift to support this strategic goal is implied, with new A3-specific roles created at all levels of the organisation.

Even to stay competitive today, operators should be setting up bodies to manage policies, procedures and technology in new A3 areas, as well as a dedicated team to undertake analysis around value creation and risk minimisation of new A3 implementations. They should also consider a central governing authority such as a Chief Data Officer to drive an A3 agenda at the highest-level. Beyond this “basic skillset”, in the next 5 years operators will need to create a centralised team to support automation efforts across the organisation with more business-focussed roles responsible for ongoing development of vision and strategy. Where these teams might sit across an operator’s organisational structure is suggested in the diagram above.

See our research on analytics, automation and AI:

AI strategy: To centralise or not? (chart)

Though only a handful of telecoms operators, including Telefónica and Elisa, have an ambition to drive new revenue growth through development of their own IP in AI, all operators will need AI to permeate their internal processes to compete effectively in the long term – it is the next logical phase of cost efficiencies the industry has been pursuing over the last ten or more years. The value of AI and automation go beyond this, augmenting every decision and process to become more informed and accurate, and establishing the fundamentals for faster experimentation, which could give rise to entirely new ways of operating.

Our research outlines the roadmap to a successful AI and automation strategy, the crucial second stage of which, following the adoption of big data analytics, is to establish a centralised AI initiative. Those who have done so are especially successful in progressing from PoCs to live AI deployments, as evidenced in the chart above, based on an STL survey across more than 50 telecoms operators.

Key activities of the centralised AI unit include:

  1. Coordinating the organisation’s approach to data management
  2. Setting the AI development roadmap
  3. Building data science and software development skills and tools
  4. Evangelising the value of AI

See our research on analytics, automation and AI:

3 key AI use cases in telco networks

3 key AI use cases in telco networks

5G is going to enhance the field of AI, but AI can also play a key role in the rollout of 5G itself. This article explores the different types of use cases for AI as applied to telco networks.

3 types of network AI use cases

In previous work by STL Partners, we identified 3 core categories of AI use cases as applied to networks:

1. Fault detection, prediction and resolution
2. Network optimisation
3. Network planning and upgrades

In this article, we will run through several of the key sub-use cases in each of these categories and explain why they matter.

Fault detection, prediction and resolution

Service impairments and faults are inevitable in a telco network, so this is a critical area in which AI can play a key role. While optimising your network may be a secondary consideration, having a functioning network is the primary consideration. Faults will also often result in large costs, whether the operations and maintenance costs themselves or fines for breaching SLAs.

Given these concerns, 3 KPIs emerge for measuring fault detection, prediction and resolution:

1. Labour cost: the cost of engineers fixing the problem, as well as customer-facing roles dealing with complaints
2. Speed: how quickly the telco can identify the problem and therefore solve it; measured as mean time to repair (MTTR)
3. Customer experience

The use cases that we will run through all seek to target these 3 KPIs, whether directly or indirectly.

Root cause analysis

Historical data is essential for detecting the root cause of network faults, in fact ML models can learn to predict the likely causes of new network faults. This dada includes customer calls, type of customer premise equipment, firmware, trouble tickets and historical data on customer premise visits.

For relatively common faults, there is a richer bank of historical data and models are able to identify the cause of the fault more quickly, hence improving all 3 relevant KPIs. If using vendor-developed solutions then the learnings from other telcos’ networks can also be applied, shortening the time to find the root cause further still.

However, for less frequent service issues which are more operator specific, it is harder to fall back on vendor knowledge. Rich and deep data must exist for more sophisticated root cause detection. The more data that the models have access to, the more likely it is that they can predict when a specific faut is likely to occur before it actually does.

Fixing the problem

Automation to fix network problems has existed in the form of fixed policies written by network engineers for over a decade. Detail in the data is required to automate the recommendation of fixes without any human input. Engineers go to sites to fix a problem, logging every step they took to fix the problem, but it is essential that they also log what the actual problem was. Richness in the data is key for automation at this level.

As models learn they are able to provide more complex recommendations, detailing the fastest solution, most cost-effective, or the least impact on customer experience.

Field engineers would only now be required for the instances when the model is unable to match a solution to a problem with a high enough level of confidence. However, a level of buy-in is required from the field engineers, and it is a sensitive issue to move from human to machine solutions. As a result, many operators may see this sub-use case as one to introduce more gradually than others which are less likely to displace human workers.

Network optimisation

Network optimisation is about how to route traffic and balance workloads across the available infrastructure and assets to try and deliver the highest quality of most cost-effective service. Elisa rolled out a self-organising RAN and reported improvements in efficiency and quality of service:

  • 20% reduction in mobile network customer complaints
  • 2% improvement in CAPEX efficiency

It is possible to optimise the network manually, but with thousands of radio sites this would mean the whole team of network engineers doing nothing but re-optimising the network. Hence, the benefit of a self-optimising network which follows a similar process as for fault resolution:

  • Real-time, event-based network data highlights a service degradation, relating to a specific root cause (e.g. sharp rise in traffic in a specific area)
  • The recommendation engine consults the policy engine to find out what the operator’s pre-defined intent is in the given situation (e.g. deliver as high quality of service as possible)
  • The recommendation engine then suggests which fixed policies, also stored in the policy engine, to implement in order to meet the intent, while adhering to any constraints (e.g. boot up any assets on standby, re-route some traffic through longer paths to reduce congestion, prioritise SMS and calls over video streaming, etc.)
  • An automated system re-optimises network equipment in line with the recommendations

It is clear that this type of network is desirable, so what are some of the sub-use cases that make it achievable? The sub-use cases address both cost and quality of service, and are broken down as such:

Cost

  • Maximise use of existing network assets / recommendation engine in where to deploy base stations
  • Predict potential service level objectives violations and prioritise traffic to reduce the risk of SLA breaches, or prioritise traffic for breaches that would incur higher costs
  • Prioritise and plan firmware updates, for times that will be least disruptive for customers and dependent network equipment

Quality of service

  • Optimise distribution of traffic between cell sites and/or service disruptions, and revert to default configuration after event/peak is over
  • Service prioritisation (e.g. calls over video streaming) during demand peaks or service disruption
  • Detect hungry VNFs that are hogging shared cloud infrastructure resources (CPU, memory, storage, networking) and re-adjust service prioritisation / spin up additional resources dynamically

With 5G and a move towards more cloud-native networks, it is expected that more use cases for AI and automation will be discovered.

Network planning

Network planning had a period where it was seen as less of a priority for many operators. The operators leading the way on AI are generally Tier 1 operators who had mostly completed roll out of 4G networks and hence were less concerned with network planning. However, with the roll out of 5G this can be expected to change. In a survey conducted by Ericsson, 70% of solution providers stated that it was in network planning where they expected to see the highest returns from AI adoption.

In a previous STL Partners report, we highlighted the example of Telefónica using AI image recognition of satellite images to try and identify rural populations not currently served by their network. By comparing mapping of unknown communities with real network coverage they were able to identify underserved areas and then further use AI to deploy networks there as well. We can expect to see more uses of AI in network planning as more operators have rolled out their 5G networks

AI and the rollout of 5G

These use cases can play a role in helping operators deploy their 5G networks. We generally associate 5G with enhancing AI, but actually AI can play a key role in the deployment of 5G itself and in improving the service offering of 5G, particularly in key areas of 5G network management. This is also the case with the deployment of 4G in less developed countries. As we look to the future, AI can play an important in dynamic 5G network slicing in enabling telcos to intelligently provision network resources so that slices can be scaled up and down as needed whilst maintaining SLAs.

A more intelligent and automated approach to networks will increase margins and increase customer satisfaction. For this reason, operators should seek to place AI use cases at the forefront of their minds when deploying their 5G networks.

At STL Partners, we have extensive research and consulting work within automation, analytics and AI (collectively known as A3) both within the network domain and beyond. We will be sharing our perspective and key insights from our work very soon, so stay tuned!

Author: Matt Bamforth is a consultant at STL Partners, specialising in edge computing, 5G and private networks.

Telco Cloud insights pack

This 24-page document will provide you with a summary of our insights from our virtualisation research and consulting work:

  • Overview of Telco Cloud deployments worldwide
  • Benefits of telco cloud: state of the industry
  • Deployment approaches: implications and challenges
  • How STL Partners can support you

Just click on the button below to request your free pack.

 

Read more about Telco Cloud, Cloud Native, NFV & SDN

Research

NFV goes mainstream: How cloud-native is contributing to growth

The number of deployments of NFV and SDN continues to grow – but while some markets are moving on to the next phase, others are just getting going.

Read more

Research

Open RAN: What should telcos do?

Alongside the roll-out of 5G cores and radios, the Radio Access Network (RAN) is evolving to a more open, virtualised and distributed architecture. What are the opportunities and risks for telcos?

Register now

Research

Telco Cloud: Why it hasn’t delivered, and what must change for 5G

Telco cloud made big promises for the transformation of telecoms. It is a fundamental enabler for 5G and the exciting opportunities ahead. Why hasn’t it delivered yet – and what needs to change?

Read more