- Comcast is combining artificial intelligence with its distributed access architecture to detect and locate outages.
- Comcast is beginning to leverage edge computing platforms and the real-time telemetry data they generate to help the company detect and pinpoint outages.
- The advantages of edge computing really come together so that Comcast now has the ability to improve the reliability of fiber, coaxial cable and, more recently, power supplies.
Combine AI with distributed access architecture
Comcast is combining artificial intelligence with its distributed access architecture to detect and locate outages. “This is a step function for ultimate reliability,” said Elad Nafshi, Chief Network Officer.
Symmetrical multi-gigabit speeds, low latency and enhanced security are one of the pillars of the cable industry’s so-called “10G” plan. Another key component of 10G is network and service reliability. Comcast is starting to push it into the spotlight as the carrier introduces Distributed Access Architecture (DAA), Virtual Cable Modem Terminal Systems (vCMTS) and an emerging lineup of full-duplex (FDX) amplifiers.
In addition, a new line of DOCSIS 4.0 chips for nodes, cable modems and amplifiers, launched in collaboration with Broadcom and Comcast, will also play an important role in the future, with artificial intelligence (AI) embedded in these chips. Comcast began leveraging the edge computing platform and the real-time telemetry data it generated to help the company detect and pinpoint outages, and then quickly take appropriate steps to keep the network online in the event of a commercial outage.
Also read: African internet outage was caused by subsea cable break
Also read: Disney to acquire Comcast’s 33% stake in Hulu for $8.61 billion
Global Outage Localisation & Detection
The program, known internally as Global Outage Localisation and Detection (GOLD), is beginning to deploy and will expand as Comcast moves forward with DAA and other elements of its network upgrade, Nafshi said. GOLD is enabled by an edge computing platform powered primarily by Comcast’s vCMTS and DAA platforms, and will be enhanced by the company’s future deployment of FDX amplifiers, he said.
As Comcast begins rolling out FDX amplifiers, the edge computing platform’s AI capabilities will extend to within a few feet of a customer’s home, providing Comcast with a detailed and localised set of AI pattern detection and actionable data.
When the digital node switches to a backup battery, the telemetry data alerts Comcast because access to commercial power has been cut off. In turn, the AI in the system can also correlate and determine if some homes in the area are equipped with backup generators.
When all of this data is aggregated and analysed, Comcast can understand the alert it received and quickly determine that it is facing a commercial outage and take appropriate action.
The data can also be applied to the continuous training of underlying AI systems. “This is where edge computing allows you to tie all of this together in a way that’s never been possible before,” Nafshi said. “It’s a step function of ultimate reliability. It allows us to optimise the customer experience to a degree that I don’t think we can control.
