Ciena's Blue Planet uses AI, machine learning to prevent up to 95% of network outages

Ciena's Blue Planet is leveraging machine learning, automation and artificial intelligence (AI) to proactively prevent up to 95% of service providers' network outages.

Blue Planet, which is a division of Ciena, announced a new software platform, Proactive Network Operations (PNO), that uses closed-loop automation to find the root causes of potential outages and then prescribe the best actions to resolve the issues.

RELATED: Ciena sends Blue Planet out on its own

With PNO in place, Blue Planet claims it can prevent network outages based on analysis of Ethernet and optical signal loss anomalies.

"You can think of it (PNO) as a solution that provides AI-assisted operations to allow our customers to move along the journey to what Ciena calls the adaptive network, or the industry more generally refers to as closed-loop automation," said Blue Planet's Kevin Wade, senior director and product marketing team leader. "We're able to look at the data coming from the network, glean these different indicators of failure, and then the machine learning algorithm recognizes that this is a high likelihood of a possible outage. Then it proactively describes fixes that can be implemented using our automation software. "

The PNO software also works with other vendors' hardware, although Wade declined to name them.

While Colt Technology Services first trialed elements of PNO last year, it won't be available until the third quarter of this year. Wade said other service providers were trialing PNO as well.

"It is still ongoing as a field trial with Colt, and it's in their production network," Wade said. "So it's deployed, but not yet part of their existing operational network. We added some additional capabilities above and beyond what Colt was testing around the recommended fixes or prescribed remedial actions."

Colt's Mirko Voltolini, head of network on demand, said in a previous interview with FierceTelecom that his company was using AI and machine learning to make the network more self-healing and more self-sufficient en route to a fully autonomous network down the road. 

Some service providers may not be ready to hand over the network keys for a fully autonomous network that uses AI, machine learning and automation.

"Our high-level opinion is that we don't think operators really want fully autonomous networks," Wade said. "They want a compromise where the operator, a human, is still largely in control of setting the polices."

Blue Planet's Mina Paik, director of portfolio marketing, said operators could set the level of automation that they want when using the PNO platform. With the machine-learning algorithm, the platform continues to adapt to what each operator wants.

"The solution is obviously designed to improve its accuracy," she said. "So it's predicting and prescribing over time. It's up to the operator to control that level of autonomy that they want to give it in the network. Obviously, the more you can do this the more adaptive, self-healing, and self-optimizing it becomes. That's the vision that we're moving towards."

Blue Planet is supporting PNO through a professional services team to help service providers tailor the program to their specific needs and networks.

As part of the PNO offering, Blue Planet also created an opex savings calculator to gauge the potential savings. Blue Planet estimates that service providers on average can save up to 38% per year on "trouble-to-resolve" opex.

"We're actually very excited about the savings calculator," Paik said. "When we modeled it, we used a very conservative approach. That 38% opex savings didn't take into consideration that the savings are also going to come from being able to eliminate the SLA penalties that are associated with outages as well."