Blue Planet’s new configuration management platform is puzzle-piece in the autonomous-networks jigsaw – which is also a part of the big telco story about reliability and trust. As telcos grow bolder in their chase of efficiency gains and developing roles in AI value chain, these reputational pillars have become the table stakes.
In sum – what to know:
Configuration drift – Network outages are reputation killers for critical telcos, says Blue Planet; most stem from manual errors, which cascade imperceptibly across fragmented tooling in complex environments.
Network autonomy – AI agents can identify the drift, enforce compliance, and automate remediation – but only as part of a broader framework spanning orchestration, assurance, analytics, and governance.
AI economics – AI is driving pressure on telcos to cut op-ex and cap-ex, which is in turn driving their boldness and innovation with AI tools for OSS modernization; they are now targeting savings of up to 30%.
Catching up, a little late here, on a conversation with Blue Planet last month about the Ciena-owned firm’s new OSS configuration platform. But the discussion goes well beyond a product recap; it is rooted in the broader narrative about intelligent networks – and how telcos make AI work for their own ends, both to drive urgent operational efficiencies and to establish a strategic role for themselves in the critical AI value chain. Even sometimes indirectly, it is as much about telcos as guardians of sovereign infrastructure, and how probabilistic AI must be harnessed to help them deliver deterministic network services for governments, enterprises, and AI systems themselves.
So it’s an anytime-retelling, really – parts of which were being told even a decade ago. “This conversation has been going for 10-plus years,” says Gabriele Di Piazza, vice president of products, data science and alliances at Blue Planet. But it is also completely different now, he says. “In the last 18 months, it has just skyrocketed.” He references the TM Forum framework for ranking autonomy in network operations, going from levels zero to five (Level 0 to Level 5) – from antiquated manual-drive systems to futuristic self-drive platforms. Some big-talking hyperscalers reckon they have achieved level-five autonomy, he says; some big (more self-effacing) telcos rate themselves “one or two”.
Di Piazza explains: “This is maybe a little controversial, but the pressure on carrier op-ex and cap-ex is much higher now, and it is driving innovation. You can save five percent or 10 percent just with procurement mechanisms and so forth. Whereas AI is cool technology, but it won’t give you anything if it doesn’t deliver a substantial saving. So we are now discussing 30 percent savings as a vision [for the telco industry], and starting to collect data points. And OSS is a major topic – because where else are you going to find those savings? Where are you going to find that service velocity, where you establish a true network as-a-service environment.” Big stuff, then, like we said.
At the end of June, Blue Planet launched a new Configuration and Change Management (CCM) platform, which adds configuration visibility, compliance monitoring, policy enforcement, and change governance capabilities to the firm’s OSS portfolio. While configuration management is hardly a new discipline, it is one of the most fragmented areas in telco operations, argues Di Piazza. Many telcos still rely on home-grown scripts, manual processes, and proprietary vendor tools that only provide visibility into portions of the network. The result is a configuration ‘drift’, where settings diverge from their baselines – which goes unnoticed, until a service breaks, and an outage occurs.
“The majority of the outages are due to a misconfiguration in some sort of a device in the network,” he says. “Which are due to manual interventions – somebody changes something, and doesn’t keep a record of it. Services and networks are interconnected; you can’t just change something over here without… a holistic way of managing the change [across the whole system].” For operators of critical ‘carrier-grade’ infrastructure, presenting themselves as government-savvy providers to highly-regulated industries, promising five-nines (99.999 percent) service reliability and total trust and flex on sovereign data management and security, any outages are a killer.
Di Piazza points to examples “in the press” of “very noisy and very large-scale dismissals of executives, or fines from governments”. It does not matter, in the end, whether they are caused by human errors or machine cock-ups – but the networks have become so complex, and the pressures on operators so pronounced, that humans cannot be trusted anyway. The AI has to work, logic goes. And it has to be closed-loop, at some point – if service uptime is to be maintained, and costs are to be managed down. “It is the table stakes. Networks are moving from connecting people to connecting machines, and to connecting agents; imagine the number of connections in the network.”
He goes on: “And 5G introduces [SLA-backed services like] slicing… And telcos [want to take] a broader role, as AI moves from training in the cloud to inferencing closer to the end user. So it is all connected; this is not just a device problem – that a device doesn’t work somewhere in the network. It is a system problem; the whole infrastructure needs to be managed holistically.” Configuration management – agent-based, Blue Planet-style – is a foundational requirement for the industry’s strategy around autonomous operations and trusted infrastructure. Its pursuit of higher-grade autonomy depends on better visibility and control over how network configurations change over time.
He comments: “Sometimes the problem is not really day-zero [design and deployment], when you make the configuration. It’s the lifecycle of it, and the drift of it – and how to manage and compare configurations, going back to compliance.” Blue Planet has embedded AI agents into the new platform via its AI Studio framework. These agents can identify configuration drift, monitor compliance, compare network states, and reduce the manual effort of audits, reviews, and remediation activities. But this configuration management is not a standalone operational task, he says; it must be a part of a broader automation loop connecting orchestration, assurance, analytics, and decision-making.
To be continued…
