April 2019 | www.tmforum.

org

DYNAMIC INVENTORY
The foundation of
intelligent automation
Author: Tim McElligott, Senior Analyst
Editor: Dawn Bushaus, Managing Editor
TM Forum’s research reports are free and can be Dynamic inventory: The foundation of intelligent automation
downloaded by registering on our website

Report author: Page 4 The big picture

Tim McElligott
Senior Analyst
[email protected]
Page 7 Section 1
Report editor:
Dawn Bushaus Evolving to dynamic inventory
Managing Editor
[email protected]
Page 10 Section 2
Chief Analyst:
Mark Newman How are CSPs using dynamic inventory?
[email protected]

Editor, Digital Content: Page 15 Section 3

Arti Mehta
[email protected] Moving toward intelligent automation
Customer Success & Operations Manager:
Ali Groves
[email protected] Page 24 Section 4
Commercial Manager, Research & Media:
Make it happen – Strategies for intelligent automation
Tim Edwards
[email protected]
Page 26 Additional feature
Global Account Director:
Carine Vandevelde Dynamic inventory: The cornerstone of intelligent automation
[email protected]

Publications Marketing Manager: Page 29 TM Forum toolkit for digital transformation

Jan Lowdon
[email protected]

Chief Marketing Officer:

Paul Wilson
[email protected]

Vice President of Marketing:

Charlotte Lewis
[email protected]

Advisor:
Andy Tiller
Executive VP, Collaboration & Innovation
[email protected]
We hope you enjoy the report and, most importantly, will find ways to use the ideas,
concepts and recommendations detailed within. You can send your feedback to the
editorial team at TM Forum via [email protected]
Report Design:
thePageDesign

Published by:
© 2019. The entire contents of this publication are protected by copyright. All rights reserved. The Forum would like to thank the sponsors and advertisers who have enabled the
TM Forum www.tmforum.org
4 Century Drive, Phone: +1 973-944-5100 publication of this fully independently researched report. The views and opinions expressed by individual authors and contributors in this publication are provided in the writers’
Parsippany, Fax: +1 973-944-5110 personal capacities and are their sole responsibility. Their publication does not imply that they represent the views or opinions of TM Forum and must neither be regarded as
NJ 07054 constituting advice on any matter whatsoever, nor be interpreted as such. The reproduction of advertisements and sponsored features in this publication does not in any way
USA ISBN: 978-1-945220-48-7 imply endorsement by TM Forum of products or services referred to therein.
The big picture

4
Full automation of networks and services is the end game for communications service providers (CSPs)
striving to compete and win in the digital marketplace. To get there, most are taking small, incremental steps
such as implementing closed control loops and machine learning for specific tasks across the business.
But before automation can become pervasive, they must transform slow, siloed, error-prone and inflexible
inventory, fulfillment, management and optimization processes into real-time, Agile processes that can
cross domains and drive a positive and consistent customer experience. This approach marks the difference
between automation for the sake of it and intelligent automation that elevates the business.

Automation is a requirement for network transformation; Are network & IT transformation simultaneous?
exponential growth of smart, connected devices and the internet
of things necessitates it. By some estimates, machine-to-
machine connections will more than double between 2017 and 2%

2022, and traffic volume will increase seven-fold. It is simply not 13% 5%
possible for manual networks to keep up with the volume and
velocity of changes required in networks supporting so many
connections and devices. 31%

To some degree, automation for activities such as routing can

be built into network functions and is being incorporated into 49%

the designs of 5G core infrastructure. However, the bulk of

automation falls to software, mainly operational support systems
(OSS), orchestration systems and policy engines – with increasing n We do not intend to virtualize n We are addressing network
help from analytics. our network or move to digital virtualization and digital operations
operations together in our current projects

Encouragingly, most CSPs recognize the importance of n We are proceeding with network n We will not start virtualizing our
virtualization but don’t see the network until we have a clear
transforming OSS and business support systems (BSS) in parallel need for digital operations strategy for digital operations
with network transformation. In a 2018 TM Forum survey, 62% of n We are virtualizing aspects of our
CSP respondents said their companies are addressing the shift to network first and will address
network virtualization and digital operations together. Even better, digital operations later

13% of them went further to say their companies refuse to begin TM Forum, 2019
network transformation before addressing operations support.

5
Indeed, many CSPs are embracing the model perfected by digital Start with the basics
natives like Amazon, which views operations software as enabling A logical starting point for transformation and automation is with
technology rather than merely as supportive of the business. As fundamental support processes, and few are more fundamental
TM Forum CEO Nik Willetts often notes: “IT is the business.” than the inventory systems that store data about physical and
virtual network resources. Inventory is part of the essential structure
Unfortunately, however, the survey found that 30% of CSP of a network that enables all the services layered on top. Dynamic
respondents are sticking with tradition and not worrying about inventory systems track and manage resources in real time using
OSS/BSS until later in their network transformation journeys. federated data. This approach is critical for intelligent automation.
This group likely will struggle to realize the full benefits of
transformation or achieve widespread automation. This report looks at the role dynamic inventory plays in making
every other process and system more accurate and therefore
easier to automate. Read it to understand:

n How inventory systems have evolved

For more about CSPs’ n What dynamic inventory is and how it can be used to enable
efforts to transform intelligent automation
networks and IT, see n How CSPs are federating data across multiple OSS and
this report domains to drive automation and improve customer experience
n The role for artificial intelligence and machine learning in
dynamic inventory and automation
Author: Mark Newman, Chief Analyst
Editor: Dawn Bushaus, Managing Editor
n Steps to take to move toward intelligent automation
Sponsored by:

inform.tmforum.org

6
Section 1

Evolving to
dynamic inventory

7
Although network architectures 20 years ago were simpler, inventory systems were no less fundamental
to the smooth operation of a communications service provider’s (CSP’s) business. However, the functional
responsibilities of an inventory system back then appear quaint in hindsight.

Consider ETSI’s view of what an inventory system should Improving inventory systems
manage, according to its 1999 Telecommunication Management Gartner Analyst Verne Anton aptly described inventory this
Network (TMN) architecture framework: way in 2003, and it’s still true today: “Inventory management
in the telecommunications world is a science. Networks are
n Installation and validation of all network elements and constantly in a state of change because events, occurring
associated equipment of the physical network at any time, change the status of the network, impact the
n Administration of all network elements and associated future topology of the network, and affect operations and
equipment of the physical network business processes. Operationally, it impacts the carrier’s
n Repairing and upgrading of all the network elements and effectiveness in managing its network and delivering
associated equipment of the physical network customer services. From a business perspective, it facilitates
n Guaranteed alignment of the inventory repository with the capital and operational expenditure control, while keeping
actual network elements and associated equipment of the customers satisfied with the delivery of services.”
physical network
Since then, inventory systems have had to evolve to deliver more
In other words, the inventory system only needed to know what capabilities, including:
assets a CSP had, where they were deployed, how many were
in reserve and perhaps what the port utilization was on each Auto-discovering network assets
network element. But even that seemingly simple job proved
impossible as manual processes and error-prone practices Near real-time reconciliation of databases
resulted in inaccuracies. Although improvements were made
and some manual processes eliminated, changes in the network Configuration management
have continued to pose challenges to the integrity of inventory
databases that are continually expected to do more. Integration with multiple operational systems

“ Inventory management in the telecommunications world

is a science. Networks are in a constant state of change... ”

8
 Feeding data to a wider array of operational and Inventory systems are not necessarily dynamic by design;
business support systems (OSS/BSS) it’s the way they interact with the network and systems
responsible for orchestrating the network (and the services
Managing physical, logical and virtual network elements running on top) that makes them dynamic. They become
dynamic when they include real-time, federated network data
Working in both virtual and cloud environments that can be accessed through standard application program
interfaces (APIs) by orchestration platforms, software-defined
Tracking millions, not thousands of assets networking controllers, analytics engines and other authorized
applications to provision, assure and enhance services.
Supporting on-demand service fulfillment and as-a-
service business models (transient resources) The dynamic inventory system maintains a single source of
truth for data. Intelligent automation happens when the data
Making inventory dynamic is analyzed and machine learning and AI are applied across the
As data increasingly drives business and operational decisions, entire process.
inventory must be as dynamic as data dictates. Inventory
systems must not only be able to ingest and respond to In the next section, we’ll look at some examples of how
external data, including analytics, they also must be the CSPs are using dynamic inventory as part of their digital
aggregator and often arbiter of the data, the judge for any transformation programs.
dispute as to the true state of the network.

“ The dynamic inventory system maintains a single source of truth

for data. Intelligent automation happens when the data is analyzed
and machine learning and AI are applied across the entire process. ”

9
Section 2

How are CSPs using

dynamic inventory?

10
Inventory becomes dynamic through data federation, standard APIs and advanced analytics, and serves as
the foundation for network transformation and automation. Most communications service providers (CSPs)
realize they need to move toward dynamic inventory, and the examples that follow illustrate some approaches.
However, they also illustrate the challenges operators face, which include competing priorities, internal
friction, embedded processes and skepticism about automation.

Use case 1: Aiming for intelligent automation accurate inventory, and recovering lost or unused assets. In
A Tier 1 multiple-system operator (MSO) in North America short, he says, OSS drives the company’s profitability.
is transforming its operational and business support systems
(OSS/BSS) to drive automation. At the same time the company While the director declined to say how much the company is
is improving its processes to support not only current network saving, other CSPs’ experiences indicate the potential savings
technology but also future technologies. The transformation is part from OSS transformation is substantial. Vodafone Group, for
of an overall push to digitize the business and become more agile. example, created an automation team in 2017 and through cost
savings derived from automation was able to increase margins
The first step for the company’s OSS team has been transitioning by 3% in 2018.
to next-generation platforms, including dynamic inventory,
that are capable of supporting real-time provisioning of virtual Federating data across systems
network functions (VNFs). Acknowledging that network drives One of the MSO’s goals has been improving the onboarding
the business, an operations software lead for the company process for VNFs to reduce the time it takes to deploy them
says that OSS is responsible for making the network scalable. and improve customer experience. Once network functions
This means automating the lifecycle of differentiated services are onboarded, automating service activation, optimization
that can be deployed across multi-vendor and multi-partner and management is easier, the director says, adding that using
environments, and automatically scaled up or down using zero- federated data and dynamic inventory helps.
touch provisioning and repeatable processes.
“Trying to ensure the data we have is accurate, we use discovery,
OSS is also responsible for keeping costs down by optimizing integration and heavy federation across all OSS and BSS systems,”
resources and configurations, monitoring utilization, ensuring he says. This has lead to many benefits (see infographic on page 12).

“ Vodafone Group has increased margins by 3% through automation. ”

11
Benefits of federated data This visibility has saved the MSO thousands of man-hours per
month in network operations.

Use case 2: Making inventory the hub

Another large MSO began its transformation two years ago by
Eliminating 2 of 5 inventory systems focusing on improving customer service. Now the company
is adding 5G network and IT transformation, which includes
improving inventory management through automation and
analytics.
Correlating data to detect performance issues
According an inventory systems manager at the MSO, auto-
Federating discovery of network assets and reconciling what is discovered
data across with the inventory database are hot topics for the OSS team.
all OSS Better order flow-through rates
But getting operations personnel and company executives to
trust the discovery results is an ongoing effort that is proving to
be a challenge, he adds.
Automation of network planning by adjusting
capacity automatically based on KPIs The inventory team performs nightly and weekly discovery
exercises, and based on the results creates discrepancy reports
to identify mismatches between the planned network and
Support for future network technologies components that are discovered. Team members know how
the planned network should look, but they are not always
TM Forum, 2019
aware of changes other teams may be implementing, which
creates discrepancies. Because of this, results aren’t considered
Dynamic inventory trustworthy enough to support automation. The manager
As a result of federating data, the MSO has been able to implement says, however, that applying more internal rules and changing
dynamic inventory, which delivers an end-to-end view of: processes so that teams can notify each other about changes
could help remedy this.
n A component’s status
n Latest orders placed for new services in any of its markets Centralizing inventory and federating data
either through sales or customer self-care As part of the transformation, the MSO has been able to
n Outstanding issues or trouble tickets consolidate five inventory systems into a single federated view.
n Performance history and customer experience However, the company keeps physical fiber inventory and

12
logical inventory separate by operating it as a dual system. Now CenturyLink is delivering on four pillars of transformation:

At the same time the inventory team was consolidating Increasing visibility into its network – neither one of
inventory systems, it was also incorporating new lines of the merged companies had visibility into the other’s
business into the central system. Now the company is working network; for example, even with physical visibility,
on stitching the data together to provide more insights across the language used by Level 3 to describe and
its network. define assets did not easily translate into something
CenturyLink’s inventory systems could understand.
Data federation has helped the MSO correlate data from
disparate systems, but federation is limited in its ability to Delivering business-owned automation – the
overcome internal barriers if one group within the company is merged company has worked to get buy-in from the
not sharing data other groups need. C-suite to get behind automation and take it a step
further by empowering the business units to take
The goal is to “make inventory be the hub and achieve ownership of their own automation.
integration between other systems such as trouble ticketing
and alarm monitoring, and get people to leverage that data and Network reliability – CenturyLink is auditing the 35
ultimately to trust the data and see the power of inventory,” the million lines of configuration code across its network
manager says. to look for anomalies and mitigate their impact
on customers; the company also added enhanced
Mergers drive transformation, integration and automation monitoring capabilities that leverage predictive analytics
Nothing gets a transformation jumpstarted like the merger of and for change management and proactive alarms.
two large CSPs with disparate networks, IT architectures and
cultures. Indeed, CenturyLink’s transformation after acquiring Transformation of skills – in anticipation of the impact
Level 3 Communications in a $34 billion deal that closed in on its employee base due to automation, CenturyLink
November 2017, has been a long road. One year into merging is training operations staff to leverage tools such as
the companies’ operations, CenturyLink President and CEO robotic process automation and data analytics.
Jeff Storey said during an earnings call that transformation
goes beyond mere integration to increase effectiveness of the To ease the burden of consolidating systems and
business, reduce costs and improve customer experience. migrating services, as well as quicken the pace of
integration, CenturyLink used federated inventory to
link multiple databases and subsystems into a unified
view that correlates issues that cross domains.

13
 The company opted to use internal developers to project: The Catalyst was designed to alleviate concern among
build automation capabilities into processes driven network operators about the fate – in a virtualized network –
by inventory, such as provisioning, activation, of a fallback tool for restoration that relied on a configuration
planning and service assurance. It remains to be known as “the last known good state” of the network.
seen, however, if the company can continue to keep
that expertise in house after reducing its workforce The approach was to embed a simple, low-cost microservice
by 12% in 2018 with 6,000 jobs eliminated. into each VNF that could report configuration changes into a
centralized inventory system. The concept keeps the state of the
Searching for a true network state virtual network predictable by enabling automatic recovery when
Verizon has long been a champion of network integrity and things inevitably fail. Some auto-discovery engines do not detect
stability, and twice in recent years has championed TM Forum changes until the process is run or scheduled and may not be
Catalyst proofs of concept that demonstrate inventory- able to fall back to the last good state of the network. And not all
and configuration-related applications aimed at improving inventory systems maintain a configuration database.
performance. In 2016, it was the Project Lily Catalyst, which
used network functions virtualization (NFV) and software- In the real world, such a solution might be centrally located in
defined networking (SDN) to simplify logical inventory by duplicated memory, rather than within the VNF, so if there is a
modelling it through abstractions at multiple layers in the catastrophic event, the VNF can be instantiated with the right
network and OSS/BSS environment. In the process the configuration without disrupting service. The Catalyst resulted
company explored the concept of creating distributed sources in an enhanced TM Forum Open API for resource configuration
of abstracted truth rather than a single source of truth for and activation to support VNF activation and incremental
inventory – a concept that hasn’t been widely adopted but configuration change notifications.
which has potential in mature cloud environments.

In 2018, Verizon sponsored the Golden Configuration Catalyst,

Watch the team
which took a page from the microservices playbook to deliver
discuss the Golden
a proof of concept for automated recovery from catastrophic
Configuration Catalyst
failure. Whereas Project Lily relied on NFV in part to solve
a problem, NFV was an indirect cause of the problem in this

14
Section 3

Moving toward
intelligent automation

15
While many communications service providers (CSPs) are taking steps to federate data and centralize
inventory systems, intelligent automation goes a step further. Orchestration platforms, software-defined
networking controllers, analytics engines and other applications must be able to access dynamic inventory
systems through standard APIs and analytics, while artificial intelligence and machine learning must be applied
across the entire process.

Current IT operations are too siloed and rely on too many introducing automation into operations is a desire to introduce
manual processes to keep pace with the new kinds of services new services that require faster and more complicated
CSPs hope to deliver, such as 5G-based internet of things responses from operations than manual processes can provide
(IoT) services. In a 2018 TM Forum survey of 65 executives (see graphic). More than 20% said new technology such as
from 37 CSPs, 42% of respondents said the primary driver for network virtualization is the primary driver.

Drivers for introducing automation into operations

n We are introducing new technology n We would like to introduce new

3% (e.g. virtualization) that necessitates services that require faster and
operations at speeds not feasible more complicated responses
20.5% for manual processes from operations (e.g. on-demand
services) than manual processes
n We need to change processes or
can provide
settings (e.g provisioning) more
42% often and more quickly than n Other
manual processes allow

n We have new process requirements

28% that are more complicated and
change more often (e.g partner
onboarding for IoT) than manual
6.5%
processes can cope with
TM Forum, 2019

“ Current IT operations are too siloed and rely on too many manual
processes to keep pace with the new kinds of services CSPs hope to deliver. ”

16
Intent-based management In many cases, these components will be in someone else’s
CSPs envision using flexible, multi-tenant platforms to domain – another CSP’s, a supplier’s, a platform partner’s or an
manage and monetize 5G services with very different network enterprise customer’s.
requirements. For example, IoT sensor networks require relatively
little bandwidth to transmit small amounts of data frequently, CSPs must agree to use the same information and data
whereas an application like remote surgery requires a huge models along with standard APIs so that orchestrators and
amount of bandwidth, extremely low latency, and guaranteed other components in different domains can communicate.
availability and reliability. This, combined with intent-based management, is how service
providers will automate service provisioning and management end
Dynamic inventory and end-to-end orchestration are key to end – the intent-based management abstracts the complexity
components in this platform strategy: Multiple orchestration of the network at a high level, then uses a customer’s intent along
systems will communicate with each other through standard with assurance, analytics and policy to manage it. TM Forum’s
APIs, such as TM Forum’s Open APIs, and with other network Open Digital Architecture provides a functional architecture CSPs
and infrastructure components to deliver and assure services. can use for end-to-end management (see page 18).

“ CSPs must agree to use the same information and data

models along with standard APIs so that orchestrators and
other components in different domains can communicate. ”

17
ODA reimagines OSS/BSS for 5G and beyond

To deliver the kinds of customer-controlled 5G services Another award-winning, multi-phased Catalyst illustrating
CSPs envision, they must reimagine and radically transform end-to-end management for inter-carrier services such as
OSS/BSS. Indeed, another recent TM Forum survey found Ethernet, software-defined wide-area networks and network
that 72% of CSP respondents believe 5G revenue growth is as a service has been key to demonstrating the layers of
dependent on transformation of OSS and BSS. ODA, from customer management to network resource
management, and how the Open APIs can be adapted and
TM Forum members are exploring how to do this through the used at each layer.
ODA, which provides a functional architecture that operators
can use for end-to-end management and orchestration of 5G The demonstrations have included constructing new product
and other digital services. Launched as a collaboration project offers based on a new network service and seamless
in February 2018, ODA provides a common operations and IT coordination of off-net service orders between CSPs. The
management ‘blueprint’. It combines proven cloud-computing customer-facing BSS communicates with OSS through Open
best practices with TM Forum’s work on zero-touch APIs that could be reused for any new serviced created
orchestration operations and management; digital ecosystem afterwards. Re-use is a key component of automation and
management; data analytics; AI and a suite of more than 50 lowers the cost of operations. The next phase of the project
Open APIs. will be demonstrated at Digital Transformation World in May.

Using learnings from several 5G Catalyst proofs of concept,

the ODA team is developing an implementation framework
Watch this video to
that can support 5G use cases. This work is detailed in a
learn more about
document called User guide for network slice management
the Catalyst
(GB999).

18
Intelligence = awareness self-interests – but it gives rise to legitimate concerns for CSPs
Intelligent management systems for next-generation networks looking to automate processes using closed-loop systems, where
need to be ‘aware’ at many levels: data is analyzed to figure out how the network can be optimized
and then changes are made autonomously, without any human
n Network-aware to maintain real-time and persistent awareness intervention. Examples of this might be automating the process
of the overall state of the network, which includes the state, for ordering services through a customer portal or automatically
location and interconnectedness of both physical and virtual adjusting radio configuration parameters to optimize a radio
network components access network (RAN.)
n Service-aware in order to optimize the network automatically
depending on the type of service running over it At some point, documenting the basis on which intelligent
n Intent-aware so they can decipher the policies that formulate systems make decisions and take corrective action will not be
the intended performance characteristics of a network or possible – or practical – at which point CSPs fear losing the
service and provide configuration management systems with institutional knowledge necessary to maintain control over their
the data necessary to make adjustments and match intent with own networks. Concerns such as this also drive operations
actual performance personnel to be extra-cautious when automating processes
around inventory and other support systems that affect
Can awareness go too far? visibility into the true state of the network. The faster and more
In AI circles there is speculative discussion about systems one voluminous network data becomes, and the more reliant CSPs
day becoming self-aware. It is science fiction currently – the become on leveraging intelligent systems to make sense of it, the
idea that an AI system might make decisions based on its own more certain they must be about the accuracy of the processes.

“ With ‘self-aware’ systems, CSPs fear losing the institutional

knowledge necessary to maintain control over their own networks. ”

19
As the operations lead interviewed for this report says about the Once CSPs have operationalized a process or tool properly,
potential for inadvertently automating bad processes, “We don’t they have incentive to move quickly. In our AIOps survey, more
want to automate something we haven’t really shaken out. We than 60% of CSP respondents said that a primary driver for
like to spend some time getting something to work before we using AI in operations is to reduce operational expense (OpEx)
automate it. We want to have it operationalized properly.” by using automation and closed-loop systems. The only driver
to score higher was delivering better customer experience.

Primary drivers for AI

Delivering a better customer experience 77%

Reducing OpEx in the organization by driving
62%
automation and closed-loop systems
Predicting performance and prevention of outages and failures 45%

Improving efficiency, agility and transparency 32%

Upselling new products to existing customers 28%

Creating new products and services
25%
(e.g. automatic voice translation as a service)
Supporting decision-making in operations teams 18%

Creating new services or service flexibility (e.g on-demand) 15%

Identifying market trends and
15%
assisting in investment decisions
Improving revenue assurance 14%

Identifying services and timing for decommissioning 2%

TM Forum, 2019

20
Leveraging AI from suppliers to provide these capabilities, but by 2020 nearly
CSPs expect to increase their use of AI for inventory- a third expect to be building AI analytics capabilities into their
supported solutions such as network planning, optimization own products and services, and nearly as many said they plan
and management in the next two years (see graphics). For the to bring AI experience in-house.
most part, operators are currently relying on outside assistance

CSPs’ use of AI for network management

2018 2020

11% 17%
27%
9%

43%

24%

38% 33%

n No activity underway n Building into products and services n No activity underway n Building into products and services

n Involved in PoCs n Building in-house expertise and n Involved in PoCs n Building in-house expertise and
adding to roadmaps adding to roadmaps

TM Forum, 2019

“ CSPs expect to increase their use of AI for inventory-supported solutions such

as network planning, optimization and management in the next two years. ”

21
The barriers to dynamic inventory and intelligent automation CSPs interviewed for this report acknowledge that silos
are largely not technical, nor are they due to a lack of available across the business are a problem, and they point to other
solutions. Granted, work remains on creating and agreeing on internal obstacles centering around systems integration and
standards for exposing and accessing data streams, but the incompatible processes such as workflow, order management
primary issue, according to most survey respondents, is inability and change management. CSPs also lack confidence in the
to access data because of operational silos. integrity of data or the rules created to make decisions based
upon it. Therefore, they are reluctant to rely on the automation
Chief barriers to exposing operational data process and relinquish control to automated systems. The MSO
inventory manager, for example, said it is likely that the broader
operations teams simply need to be exposed to the automation
1.5%
5% process and its results in order to build confidence.

21% Almost a quarter of respondents cited work on standard

APIs for exposing data as a significant obstacle, although
44.5%
some progress has been made in this area. To date, 45 of the
world’s leading service providers and technology ecosystem
participants have signed TM Forum’s Open API Manifesto
29% publicly demonstrating their endorsement of TM Forum’s suite
of Open APIs.

n Inability to access the data needed n The lack of relevant APIs for
across operational silos exposing data streams
n The lack of clear standards for n Lack of provisions for/concerns
exposing data streams about security

n Other
TM Forum, 2018

“ The barriers to dynamic inventory and intelligent automation

are largely not technical; they are due to operational silos. ”

22
Collaborating on AIOps

TM Forum members are working on an AI reference service operations management to show a more comprehensive,
architecture, which includes developing standard interfaces end-to-end view of how AI can benefit CSPs.
for operational AI models and standard component structures
that define the architectural components used to develop “It is mandatory to use new technology as much as possible for
and manage AI models, such as feature databases, model improving the service that we offer to our customers,” says Massimo
databases, deployment servers and registration databases. The Banzi, Standards Manager, Telecom Italia, one of the champions of
idea is to show how real examples of AI components could be the project. “We need to be flexible and we need to anticipate the
usefully developed and managed. need of the customer. Hence, artificial intelligence – which comes
from the continuous evaluation of data collected from all possible
A new information guide entitled Service management standards sources. This means collecting data from several categories of
for AI discusses the motivation and drivers for AI management customer and collecting data from the network to understand
and examines the lifecycle of an AI model from development when there are performance issues, then tackling them by
through to decommissioning a service, presenting best practices dynamically redirecting, adding and scaling network resources, etc.”
in each stage of the lifecycle. It also defines a high-level model
for AI management as well as a reference architecture.
Watch this video to
Traditional software development lifecycle methodologies often learn more about
fall short when it comes to managing AI models. This means the Catalyst
it is critical that CSPs and their suppliers develop an agreed-
upon framework now for managing the lifecycle of AI models in
business applications. Trend Analysis Report

AI AND ITS PIVOTAL

ROLE IN TRANSFORMING
OPERATIONS
A multi-phased TM Forum Catalyst proof of concept called AI for Author: Catherine Haslam, Senior Analyst

For more about CSPs’

Editor: Dawn Bushaus, Managing Editor

IT & network operations, which will be demonstrated at Digital

December 2018

efforts to transform
Transformation World in May, is looking at how to address
networks and IT, see
many AIOps challenges. In the first phase of the project, the
this report
team explored five use cases around how to improve customer
experience and efficiency by optimizing IT processes and
operations. This phase extends AIOps to additional areas of CSPs’

23
inform.tmforum.org
Section 4

Make it happen –
Strategies for
intelligent automation

24
The first step toward any goal is deciding to go for it. CSPs should stop wavering in their commitment to
leverage intelligence and automation as they will be unable to manage their businesses competitively or
profitably without them. Nor will they be able to meet customers’ expectations around speed to market,
flexibility, agility, programmability, control and responsiveness. Following are steps CSPs must take to
automate intelligently:

Get C-suite support Federate data so it can be shared

Leaders of IT teams within CSPs should work to Legacy OSS/BSS applications were not designed
get support from executives to invest in improving to provide real-time data or respond accordingly,
foundational operational support systems (OSS), but these systems and databases contain critical
especially inventory. It is important to get the leaders information that must be synchronized with real-time
of the business to recognize the value in network data in next-generation systems. Invest in federating
data and the added value of sharing it across the data from legacy systems and centralizing it in a
business and with partners. dynamic inventory platform or hub where it can be
shared with other systems.
Start with the basics
Resist moving forward and inadvertently automating Apply analytics
bad processes. Create or refine fundamental Whether through skills developed in-house or in
processes associated with inventory, fulfillment, collaboration with data-science partners, CSPs
catalog and assurance so that they reliably achieve should apply analytics and artificial intelligence (AI)
automated, error-free, flow-through results today to all data generated by network infrastructure, even
and as they apply to any next-generation functions outside plant. To learn more about AIOps, consider
and services. joining a TM Forum Catalyst proof of concept or
joining the AI collaboration community. For more
Focus on closing the loop information, please contact Aaron Boasman-Patel via
In networking, ‘closing the loop’ means collecting and [email protected].
analyzing data to figure out how the network can
be optimized and then implementing those changes
automatically. As incrementally as necessary, but as
quickly as possible, OSS/BSS teams should address
each process or microprocess with the goal of closing
more loops and removing manual intervention.

25
SPONSORED FEATURE

Reaching new heights founded on a dynamic inventory cornerstone

In building construction, the cornerstone is the critical n Trouble to resolve – Proactive monitoring of the service to When it comes to capacity planning, historical metrics
element to ensure a reliable, long-standing structure. If the handle events and faults and restore services to a normal state typically are painstakingly, manually analyzed so sufficient
cornerstone is not solid, the entire structure is shaky and network capacity can be built out at the needed locations
will not be able to withstand harsh conditions. To fulfill their business functions, operations personnel to satisfy customer demand. The proliferation of OSS
rely on IT systems that are underpinned by inventory silos leads service providers to over-engineer to minimize
In the world of networking, what serves as the data, yet that data is fragmented, spread across multiple service order failure and loss of revenue.
cornerstone? How can service providers embark on digital inventory systems. For example, there is network
transformation and network automation projects if they are inventory data (such as device or port), geolocation Due to these shortcomings, over time, business processes
not aware of the resources underpinning their networks? To inventory data (for example, if a router is on the third floor have become riddled with errors, ultimately driving higher
effectively manage their networks at the most basic level, of a building at a specific address), and passive equipment operational costs and a poor QoE for the end-customer.
service providers need an inventory system that accurately data (such as patch panels or cross-connects). There is
presents all available resources, physical and virtual, end to also data associated with the state of subscriber services. Blueprint for success – Blue Planet® Intelligent Automation
end. Only after laying this foundation can operators build This data exists for multiple domains (such as access, Blue Planet enables service providers to bridge IT and
automation capabilities, block by block: next comes service metro, core) and network layers (such as optical, Ethernet, network operational silos and re-evaluate and optimize
orchestration within a domain, and across multiple domains, IP). It is not uncommon for large providers to have dozens the critical business processes, leveraging the key
as well as route optimization and assurance. Furthermore, of different inventory systems, either vendor-supplied or elements of intelligent automation—dynamic inventory,
the accuracy and precision of business processes can internally developed. Service providers have steadily built systems federation, analytics, policy management, multi-
be augmented with advanced analytics powered by AI overlay network infrastructure and Operational Support domain orchestration, network and service topology and
techniques. With these analytics building blocks, network Systems (OSSs) to meet each new service demand. visualization, and route optimization and assurance.
automation can reach new heights. Predictive analytics
serve to drive informed, prescriptive actions that enable Moreover, these legacy inventory systems are static; they
networks to adapt rapidly to changing market demands, not informed by real-time network information. Today,
ultimately enhancing the end-customer experience. as virtualization technologies are introduced alongside
traditional, hardware-based infrastructure, OSSs must become
Legacy operational systems have irreparable cracks more in sync with the live state of the managed networks. A
Within a service provider, three interconnected network and static, inaccurate view of resources leads to a breakdown of
IT processes are integral to business operations: the mission-critical processes that span IT and network.

n Plan to build – Creation of the service-ready network that Multiple assurance systems also exist per domain and per
is able to support new services, including physical, logical, layer. These present real-time performance metrics, yet
and virtual infrastructure are not correlated, thus making it difficult to determine
n Order to service – Taking the customer request and converting how issues in one domain or layer impact end-to-end Figure 1. Blue Planet solutions automate critical business processes
it into service via the network resources and applications services, and to troubleshoot the root cause of issues. that span service providers’ IT and network operations
SPONSORED FEATURE

Reaching new heights founded on a dynamic inventory cornerstone

Each service provider line of business has a product Ultimately, Blue Planet enables closed-loop automation. Order
Mgmt
Fault
Monitoring
catalog of services that span any mix of network Network state and performance telemetry generated Web-based GUI/ CPEs
domains; therefore, operations personnel need to access by instrumented physical and virtual infrastructure SDWAN
Portal
REST APIs
Trouble
multiple, per-domain inventory and monitoring systems help derive business insights, which in turn help Request View Ticketing

to complete service planning, fulfillment, and assurance define and refine software control policies. Ongoing
Contact
Mgmt Physical
Inv

tasks. Blue Planet Intelligent Automation synchronizes analytics trigger policy-guided actions back into Blue Planet
Federation Layer SDWAN
Service
per-domain information and correlates real-time network the programmable network, enabling automation of
Perf Mgmt
Assurance

data to present a single, accurate view of network network planning, service fulfillment, and assurance Customer
resources. This gives operators the ability to create processes. This open, data-driven approach optimally Portal
Reporting
Schema
customer services that combine any domain resource aligns and accelerates the critical business processes Alarm
Blue Planet Database
Mgmt
and manage the full lifecycle of operations for that that span IT and network operations, thus delivering JDBC Rest APIs Golden Gate Replication
service—including troubleshooting and SLA monitoring— faster time to revenue, reduced costs, and a better end-
across multiple vendors and hybrid physical and virtual customer experience. Inventory 1 Inventory 2
Network
Alarms
Order
Management
Performance
Management
...

network infrastructure. By abstracting the complexity of

the underlying multi-domain infrastructure, Blue Planet Laying the dynamic inventory foundation
Hybrid
greatly accelerates and simplifies business operations that The first step to holistic automation involves federation Network
Orchestration
4G Backhaul
Network
Virtual Ent.
Services
Ciena
Packet-Optical

rely on both IT systems and the network. of multiple per-domain, per-layer inventory systems, VNF VNF VNF

along with reconciliation with real-time network

data to create a single, dynamic view of all network Figure 3. Blue Planet’s unified inventory underpins business operations
resources. This approach allows service providers to
leave their existing system investments in place while Layering automation building blocks
introducing a ‘single pane of glass’ with a user-friendly, With the dynamic inventory foundation in place, service
next-generation interface. Innovative visualization of providers can raise their automation capabilities with
end-to-end network and service views dramatically multi-domain and NFV orchestration. This software
improves the ability to troubleshoot network issues and enables them to dynamically create, deploy, and
accelerates the network planning and design processes. automate the end-to-end delivery of services across
Furthermore, many tasks that were extremely tedious both physical and virtual networks. It provides a
and error-prone can now be automated due to Blue scalable and open software abstraction layer that
Planet’s open APIs, which allow for simple integration to enables seamless service automation across any mix
Figure 2. Blue Planet solutions automate critical business processes adjacent OSSs. of vendors and network layers. Leveraging an agile,
that span service providers’ IT and network operations
SPONSORED FEATURE

Reaching new heights founded on a dynamic inventory cornerstone

DevOps-style approach toward resource onboarding and trigger automated actions back into the programmable smarter and more agile so they can adapt to changing
service development, it also allows service providers to network infrastructure. In this way, the network can market dynamics more quickly and remain relevant to
easily and rapidly incorporate new networking resources adapt to changing demands and resolve problems their markets.
and bring innovative new services to market. before they even occur. Examples include adjustment
of bandwidth allocation, rerouting of a service path, Blue Planet’s Intelligent Automation portfolio
Additionally, Blue Planet route optimization and instantiation of virtual functions, or auto-scaling. And enables service providers to succeed in their digital
assurance capabilities uniquely combine routing, traffic, the cycle continues indefinitely—an adaptive network transformations by bringing together best-in-class
and performance analytics for real-time, path-aware continually generates telemetry for ongoing analytics to dynamic, federated inventory, analytics, multi-domain
operational monitoring of IP/MPLS networks, and back- scale and self-optimize. orchestration, and route optimization and assurance
in-time forensics to troubleshoot problems that can cause capabilities. Built upon an open, microservices-
service disruptions. Interactive modeling helps engineers based infrastructure, these capabilities work together
optimize their networks by helping to predict the impact to achieve a closed-loop approach to intelligently
of changes, simulate new workloads for capacity planning, automating mission-critical business processes that
and test failure scenarios. span IT and the network. This approach is unique
in the industry. Blue Planet offers expertise in both
Designing windows into the network networking and IT automation, and highly specialized
Instrumented network infrastructure generates a wealth professional services, to deliver a comprehensive
of network data that informs IT systems with real-time solution. Furthermore, Blue Planet provides an open,
information and can be used for diagnostic purposes. Figure 4. Analytics methodologies offer increasing degrees of programmable platform that can be leveraged for
Advanced analytics goes beyond simply analyzing business value continued enhancements using DevOps methodologies.
patterns and trends and taking reactive actions—it
enables proactive operations. Using AI and machine Built to last, built to scale With Blue Planet, service providers can evolve from
learning methodologies, operators gain the ability to Service providers are striving to outpace their the constraints of their legacy environments to an agile
more accurately predict potential network problems and competition to meet the ever-increasing expectations operational approach based on greater insight and
anticipate trends. Mountains of collected data can be from end-customers for differentiated and on-demand synchronization, and the automation of IT and network
turned into actionable insights that drive the creation and services, high bandwidth, and high QoE. To succeed, processes. The result is closed-loop automation that
fine-tuning of data-driven business policies. These intent- providers know they need to undergo a digital maximizes operational flexibility to deliver a better customer
based policies are used by the software policy engine to transformation—making their IT and network operations experience while driving top-line revenue growth.

“ This open, data-driven approach optimally aligns and accelerates the critical
business processes that span IT and network operations, thus delivering
faster time to revenue, reduced costs, and a better end-customer experience. ”
TM Forum toolkit for
digital transformation

29
TOOLKIT

Agile & Virtualized Open & Partner Effectively Customer Centricity

TM Forum Digital Maturity Model Open APIs Customer Experience Implementation Suite
The DMM is a ‘living’ maturity model and set of TM Forum offers more than 50 APIs to manage This set of tools consists of a guidebook, hundreds of
metrics to help companies measure their true digital services end to end and throughout their lifecycle in a metrics, a maturity model, lifecycle model, ROI model
maturity. Members can access a guidebook as well as multipartner environment. and more than 54 implementation use cases.
an Excel spreadsheet containing the actual model. It is
also available as an iOS app.
Digital Trust Challenges and Opportunities Standard Big Data Analytics Solution Suite
This technical report outlines the key concepts of digital This set of tools includes a big data reference model,
Agile OSS/BSS Toolkit trust and identifies the top seven digital trust challenges. a guidebook containing more than 65 use cases and
This toolkit includes a complete blueprint for a 1700+ pre-defined metrics.
platform for managing a multi-vendor hybrid/NFV
infrastructure, which includes open APIs, information Monetizing the Internet of Everything Guide
models, best practices and deployment guides. This information guide describes a standardized 360 Degree View of a Customer
approach and a monetization template for new, This guidebook offers a 360-degree view of a
innovative services. customer and explains how to put customers at the
Open Digital Architecture center of considerations and actions.
Developed collaboratively by the world’s largest telecom
operators and their partners, the ODA provides a
common operations and IT management ‘blueprint’.
It combines proven cloud-computing best practices
with TM Forum’s work on zero-touch orchestration
operations and management; digital ecosystem
management; data analytics; AI and Open APIs.

30
 TOOLKIT

TM Forum research reports

March 2019 | www.tmforum.org
Trend Analysis Report Trend Analysis Report

AI AND ITS PIVOTAL REDEFINING THE CTIO

CT
ROLE IN TRANSFORMING ESSENTIALS FOR THE DIGITAL AGE

OPERATIONS Author: Catherine Haslam, Senior Analyst

Editor: Dawn Bushaus, Managing Editor

TRANSFORM SERVICE
November 2018
Author: Catherine Haslam, Senior Analyst
Editor: Dawn Bushaus, Managing Editor

CTIO
R E D E
December 2018

ACTIVATION FOR
WITH REAL-TIME
F I N I

IO
DATA ANALYTICS
Author: Sandra O’Boyle, Independent Analyst
Editor: Dawn Bushaus, Managing Editor

N G T H E
C T I O
inform.tmforum.org inform.tmforum.org

Digital Transformation Tracker 2

How to fix the

cultural divide
Author: Catherine Haslam, Senior Analyst March 2018 | inform.tmforum.org
Editor: Dawn Bushaus, Managing Editor

Author: Mark Newman, Chief Analyst: [email protected]

Editor: Dawn Bushaus, Managing Editor: [email protected]

Sponsored by:
Author: Mark Newman, Chief Analyst
Editor: Dawn Bushaus, Managing Editor

Sponsored by:

Executive Summary
TM FORUM QUICK INSIGHTS REPORT TM FORUM QUICK INSIGHTS REPORT inform.tmforum.org TM FORUM QUICK INSIGHTS REPORT TM FORUM QUICK INSIGHTS REPORT
inform.tmforum.org
inform.tmforum.org Sponsored by:

E ION
CU S TO M
GA RIV
KEEDTRIVEDALD SFORMAT RPTAY
O N
LE
LDI
A A
ER

L I F E C YC

BUI
TM Forum Open APIs:

A L I T S M WANT TO DRIVE
& CUSTOMER
LIF

T IVED DIG ONTR

I V N
Enabling a zero-integration API economy

DVA
R A CS BUSINESS BENEFITS?
L LYTIIDENDS EXPERIENCE:
E C YC

IMPROVE CUSTOMER
R
A TI A
IGITSFORM
E D A N
A DIV
VED
RI EXPERIENCE EMERGING BEST PRACTICES

DRI ALD TRAN

CU AC H
E

ROSS T
E
LE

Author: Sandra O’Boyle, Independent analyst

T STOM
Editor: Dawn Bushaus, Managing Editor

I
st
Analy
ndent tor Author: Mike Hibberd, Independent Analyst & Consultant

G
Indepe ing Edi

N
ner,

I
nag Editor: Dawn Bushaus, Managing Editor
nie Tur us, Ma

I O
st r: An
Analy Busha
Autho Dawn

D RMA
May 2018

T
ndent itor r:
Indepe g Ed Edito
lson, Managin
rol Wi us,
r: Ca Busha
Autho Dawn
r:

SFO
Edito

RAN

31
TM Forum Frameworx

32
FRAMEWORX

TM Forum Frameworx is a suite of best practices and

standards that, when adopted, enable a service-oriented, highly
automated and efficient approach to business operations.
Frameworx provides hundreds of standardized business metrics
that allow for benchmarking, as well as a suite of interfaces
and APIs that enable integration across systems and platforms.
Frameworx also includes adoption of best practices to help
companies implement and use the standards and ensure
ongoing conformance.

Frameworx has been widely adopted and proven to significantly

improve agility in IT and operations, resulting in increased
margins, lower costs and optimal customer experience.
Frameworx is created and evolved by TM Forum members who
participate in the Forum’s Collaboration Community.

8 things Frameworx can do for you: 5. Optimize business processes to deliver highly efficient,
automated operations
1. Reduce transformation risk by delivering a proven blueprint 6. Reduce integration costs and risk through standardized
for agile, efficient business operations interfaces and a common information model
2. Innovate and reduce time-to-market with streamlined end- 7. Gain independence and confidence in your procurement choices
to-end service management through conformance certification and procurement guides
3. Create, deliver and manage enterprise-grade services across 8. Gain clarity by providing a common, industry-standard language
a multi-partner ecosystem
4. Improve customer experience and retention using proven
Download latest files Get training
processes, metrics and maturity models

33
FRAMEWORX

Business Process Framework (eTOM) 3. Apply disciplined and consistent business process
The Business Process Framework (eTOM) is a comprehensive, development enterprise-wide, allowing for cross-
industry-agreed, multi-layered view of the key business processes organizational re-use
required to run an efficient, effective and agile digital enterprise. 4. Understand, design, develop and manage IT applications in
terms of business process requirements so applications will
6 things you can do with the Business Process Framework: better meet business needs
5. Create consistent and high-quality end-to-end process
1. Create a common language for use across departments, flows, eliminating gaps and duplications
systems, external partners and suppliers, reducing cost and 6. Identify opportunities for cost and performance improvement
risk of system implementation, integration and procurement through re-use of existing processes and systems
2. Adopt a standard structure, terminology and classification
scheme for business processes to simplify internal
operations and maximize opportunities to partner within
Download latest files Get training
and across industries

Information Framework (SID) 3. Speed time to market by using well-understood integration

The Information Framework (SID) provides standard definitions interfaces, eliminating the need for data translation
for all the information that flows through the enterprise and between systems
between service providers and their business partners. 4. Avoid wasting precious development time on debates with
your team, partners or vendors by adopting a widely proven,
5 things you can do with Information Framework: industry accepted, rich and extensible information model
5. Mandate conformance to the Information Framework and save
1. Reduce integration costs by adopting standards-based time and money during vendor evaluation and procurement
models and using them in applications and interfaces
2. Save hundreds of design hours by starting with a mature
framework and 1,500 entities developed and vetted by
Download latest files Get training
subject matter experts

34
FRAMEWORX

Application Framework (TAM) 3. Integrate faster and with lower costs by defining and clearly
The Application Framework (TAM) provides a common language communicating the functions provided within each application
and means of identification for buyers and suppliers across all 4. Reduce custom development costs with modular, standard
software application areas. application requirements
5. Increase automation and efficiency with standard,
5 things you can do with the Application Framework: deployable components

1. Streamline procurement by using common definitions and

language to specify and evaluate solutions
2. Document and then rationalize your application inventory
Download latest files Get training
during transformation projects or mergers and acquisitions

Open APIs dramatically improving business and IT agility, reducing the cost
TM Forum’s 50+ REST-based Open APIs have been developed and complexity of operations, and reducing integration cost,
collaboratively by communications service providers (CSPs), risk and time for the entire supply chain.
government organizations and their partners. When used
internally, the Open APIs help companies transform their IT, The Open APIs are often tested, improved and extended
increase operational agility and improve customer centricity. through TM Forum’s Catalyst Program. Catalysts are proof-
Externally they enable end-to-end seamless connectivity, of-concept projects that bring together companies large and
interoperability and portability across complex digital ecosystems. small to create innovative solutions to common challenges,
demonstrating how solutions can be achieved by leveraging key
To date, 45 of the world’s leading CSPs and technology suppliers TM Forum best practices and standards. Catalyst teams work
have signed the Open API Manifesto publicly demonstrating on the projects for four to six months before demonstrating
their endorsement of TM Forum’s Open APIs. CSPs that adopt them at TM Forum’s flagship events.
the Open APIs can position them as a preferred requirement
in their IT requests for proposal, and technology partners can
commit to using the Open APIs in relevant product applications.
Access the Open APIs Learn more
Together they can unlock many growth opportunities, including

35
FRAMEWORX

Best practices
TM Forum members have collaborated to produce an extensive
library of standards, best practices, guidebooks, technical
reports and much more covering the most important topics for
companies operating in the digital economy.

We have arranged these resources into toolkits by topic. Click

on the link below to access the full toolkits and download* all
the available resources.

*Downloads are available to employees of TM Forum member

companies. Interested in joining as a member? Click here

Access the toolkits

36
Meet the R&M team
Chief Analyst: Customer Success & Operations Manager:
Mark Newman Ali Groves
[email protected] [email protected]

Managing Editor: Publications Marketing Manager:

Dawn Bushaus Jan Lowdon
[email protected] [email protected]

Senior Analyst: Global Account Director:

Tim McElligott Carine Vandevelde
[email protected] [email protected]

Editor, Digital Content: Commercial Manager, Research & Media:

Arti Mehta Tim Edwards
[email protected] [email protected]

37
 For more information about the TM Forum Collaboration
Program, please contact Andy Tiller, Executive Vice President,
Innovation & Collaboration, via [email protected]