Operators are increasingly focusing on how autonomy translates into measurable impact, whether that is improved customer experience, faster service restoration, lower operating costs, or more efficient use of energy and resources. This is where the role of clear metrics becomes critical.

Key performance indicators remain essential for tracking operational efficiency and network behavior. Alongside these, there is a growing emphasis on key business indicators, which connect autonomous capabilities directly to outcomes such as revenue protection, customer satisfaction, and service reliability. What is emerging is a more complete picture of value, where autonomy is not assessed purely on technical sophistication, but on its ability to move the metrics that matter most to the business.

This is also an area where the industry still has work to do. One of the priorities for the industry is to help define and standardize how these outcomes are measured, so that operators can consistently evaluate their level of automation against real business impact, rather than relying on subjective or fragmented indicators.

When this alignment is in place, autonomy becomes easier to justify, easier to scale, and far more meaningful at an executive level. It shifts the conversation from capability to consequence.

Why Digital Twins Are Becoming Essential

Digital twins have been part of the conversation for years, often positioned as a future capability with broad potential. What is different now is that they are being applied in focused, practical ways that directly support autonomous operations.

In areas such as service assurance, fault management, and troubleshooting, operators are using digital twins to simulate network behavior before changes are applied, to test how intent policies will perform under real conditions, and to validate decisions without putting live services at risk. This creates something the industry has historically struggled with: confidence at scale.

As networks become more complex, it is no longer feasible for human operators to anticipate every possible outcome of every possible action. Digital twins provide a way to explore those scenarios safely, allowing systems to learn not only from live environments, but also from modelled ones. What is important here is that successful implementations are not attempting to model the entire network in one go. They are targeted, domain specific, and incremental, building trust over time rather than attempting transformation in a single step.

Building Coordinated Intelligence Across Domains

Much of the progress the industry has made so far has been within individual domains. Operators have successfully applied autonomous capabilities in areas such as fault management, transport optimization, and energy efficiency, often delivering strong results within those boundaries. These achievements matter, and they are an essential part of the journey.

But networks do not operate in isolation by domain. Customer experience, service performance, and operational efficiency are all shaped by interactions that cut across network layers, technologies, and organizational boundaries.

To move beyond partial autonomy, the industry needs to shift from single domain intelligence to cross domain orchestration, where systems can coordinate decisions across multiple layers of the network, balancing competing priorities and responding to change in a holistic way. This introduces new challenges, particularly around data consistency, policy alignment, and control across domains that may have evolved independently. It also raises the bar for architecture, requiring a level of integration and standardization that many organizations are still working towards. However, it is also where the real value of autonomy begins to unlock.