Telstra highlights data hurdles in pursuit of network autonomy

Telstra is navigating a complex transition toward autonomous network operations as surging demand from hyperscalers and neoclouds creates unprecedented traffic variability. While the company aims to automate low-risk, repeatable tasks to improve responsiveness, it maintains that human oversight remains essential for high-impact policy decisions. The primary obstacles to achieving full autonomy include poor data quality, a lack of cross-network visibility, and the need for normalized telemetry to ensure reliable automated responses.

According to Fierce-network, Telstra is currently navigating a sophisticated balance between traditional automation and full network autonomy in response to shifting industry demands. The telecommunications giant is facing a dual challenge: managing the massive scale of traffic generated by hyperscalers while meeting increasingly rigorous expectations for reliability and real-time responsiveness.

The necessity of human oversight in AI policy

Despite the push toward self-healing systems, Telstra emphasizes that autonomy must be earned through proven reliability rather than implemented overnight. The company is currently prioritizing automation for high-volume, low-risk tasks to eliminate manual friction, such as redundant checks and repetitive handoffs between teams. However, human engineers remain central to the architecture of these networks.

Regan Ireland, global head of pre-sales solutions, products and digital experiences at Telstra, noted that complex trade-offs require a human touch. He explained that while automation handles the heavy lifting of detection and recovery, humans are required for:

Defining overarching network policies

Managing unique technical exceptions

Evaluating risks associated with customer impact

Making high-level strategic decisions regarding service delivery

"Anything to do with policy, with complex exceptions…those sorts of complex trade-offs, they are something that I think there will always be a human in the loop," — Regan Ireland, global head of pre-sales solutions, products and digital experiences at Telstra.

Overcoming data normalization barriers

The transition to a more autonomous model is currently hindered by significant data quality issues. For AI to function effectively, operators require "clean telemetry" and a governed source of truth across various network boundaries. Without proper normalization, the sheer volume of data can become overwhelming, leading to a situation where critical signals are lost in a sea of noise.

By solving these visibility gaps, Telstra aims to move toward a proactive operational model. The goal is to replace standard "we will get back to you" communications with faster quotes, quicker solutions, and real-time insights into network status. Ultimately, the objective is not to create a hands-off network but one where automation removes repetitive labor while human judgment defines how the operator manages resilience and risk.

FAQ

Why does Telstra still require human oversight in network automation?

Telstra maintains that humans are necessary for high-impact policy decisions, managing unique technical exceptions, and evaluating risks associated with customer impact. While automation handles detection and recovery, complex trade-offs and overarching network policies require a human touch to ensure reliability.

What data challenges is Telstra facing regarding network autonomy?

The transition is hindered by significant data quality issues and a lack of cross-network visibility. Operators require clean telemetry and a governed source of truth because, without proper normalization, the volume of data can cause critical signals to be lost in noise.

The necessity of human oversight in AI policy

Overcoming data normalization barriers

FAQ

Fresh news on our Telegram