Zero-touch 5G automation is a marathon, not a sprint

For many network operators, the prospect of a fully automated network operations centre (NOC) is compelling: a zero touch, lights-out NOC has a lower requirement for human presence and manual interventions in day-to-day network operations, which in turn minimizes OPEX and reduces network downtime and Mean-Time-To-Resolution (MTTR). Whilst this sounds great in theory, the reality is a little more complex. Whether full automation in the NOC is an achievable goal is a matter for debate.

In green field environments, or where incumbent operators are in the process of building brand new stand-alone networks—as can be the case with some 5G roll-outs—operational automation is introduced at inception. Even in these scenarios however, the picture isn’t that clear-cut. Maintaining operational automation is a process that requires consistent on-going investment and continuous effort. Without these two elements, the automation aspect of your operations will degrade over time.

It is also difficult to transition from a manual or semi-automated legacy environment to a zero-touch network management approach when you have to upgrade existing tools and processes, whilst also having to prove that there is a valid business case to do so.

There are many practical hurdles to consider when it comes to embarking on this type of large scale, long-term NOC automation project but it can be achieved if a more pragmatic, balanced approach is taken. By automating the ‘low hanging fruits’, with a step-by-step phased approach, CSPs can demonstrate that the journey to automation in the NOC is a marathon, not a sprint.

Finding value in automation

The incentives to create a zero-touch service assurance environment are varied, but what often motivates CSPs to roll-out their automation initiatives, is an acute need to achieve an increase in overall productivity whilst not increasing OPEX costs. These types of phased projects often occur organically and are conceptualized tactically to overcome a specific challenge or issue and to achieve performance and cost-saving goals. For the NOC, this might be achieved, for example, through the reduction in the amount trouble-tickets and the required for intelligent trouble-ticket creation through transition from rules-based to machine-learning driven tools and processes. This kind of automation typifies the long-term, incremental approach to network management automation that evolves over time towards a more comprehensive zero-touch service assurance environment – where you take an existing known process and replace the manual or rules-driven approach with ML and AI.

Even if automation is not part of a larger corporate plan, individual departments can act autonomously where they apply their ingenuity and resources to tackle ‘bite-sized’ projects. From this focused vantage point, it is possible to easily identify opportunities that make processes more efficient, whilst simultaneously justifying the business case. The preponderance of smaller, more manageable projects that quickly prove ROI and demonstrate measurable gains assure the executive level that additional investments into service assurance automation generates value. Once proven in isolation, optimization opportunities are then easier to scale-up and roll-out across the entire organization.

Contrastingly, the inverse strategy to this bottom-up approach emerges when the executive management level aims to build a foundation from which an organization can continue to innovate, as part of an overall Digital Transformation strategy. With a significant budget and resources behind it, a large-scale automation project can be led by a long-term corporate vision that is achievable within a well-defined roadmap. With key mile-stones that are specified early and communicated clearly, will help gain buy-in can from all departments to ensure automation policies are permeate across all levels of the organization. For example, part of the strategic plan might be conducive to helping a telco fully automate a certain percentage of their processes by a specific date, or substantially reduce overall network downtime.

Grabbing the ‘low-hanging fruits’

Once the incentives and objectives towards automation are clearly defined, it is important to assess existing network operations from an IT perspective. A rationalization process of the current infrastructure can help to identify whether the right tools and capabilities exist to achieve the desired automation goals. Network operations teams should evaluate what tools need to be retained and modernized and judge which legacy items should be retired since they are not able to support the required level of automation. 

One option would be for operators to retain reliable solutions that deliver a high-level of functionality and efficiency whilst integrating new features and advancements in order to meet business objectives. For instance, it is possible to apply machine learning-based root cause analysis and predictive analytics to the existing service assurance platforms to achieve desired business outcomes without replacing the underlying basic event collection and processing functionalities. 

In terms of achieving full automation in the NOC, a greenfield scenario is a much more amenable place to start as it allows planning and implementation from inception, rather than having to backport automation into existing and established practices and systems. However, it is unrealistic to expect the same vision to be achieved in a manual or semi-automated environment as it is necessary to change and upgrade existing tools and processes, and at the same time justify the investment. 

For mature operators then, full NOC automation can only ever be a gradual progression, as opposed to a ‘big bang’ event. Despite this, the availability of ‘low hanging fruit’ that can be easily addressed could make progress towards this goal swift and meaningful.

A tipping-point

As operators continue to seek ways to optimize and modernize their networks, there is a clear and critical need to assess the current state of their existing service assurance operations and OSS estates to ensure that they meet the demands of current and newly deployed 5G networks.

The tipping point arrives when it is no longer viable to support advanced and dynamic network functions via manual means, regardless of the costs. At this stage, service assurance automation tools need to be readily available and mature enough to support operators with their automation goals. With said tools in place, it will be much easier for organizations to switch to the zero-touch mindset, embrace a new culture and set operators on the path to full, or near full, automation.