FROM THE INDUSTRY
Is that something you would look to? It’s a competitive marketplace already. There are many innovators out there of course, but what they are all doing is collect data from their own kind of products. Nobody is analysing and correlating data collected from across the entire network; for instance correlating the data from cable modems with the data that we can collect from our DAA nodes and smart amplifiers. There is another layer available and that’s exactly where the AI and ML come into play, because now we have tons of data, way more than ever before. Is this a potential line of business for Teleste? Yes, but it’s still early days; we need to go deeper on this topic with kind of lead customers, because some of our US customers have been doing quite a lot of in-house development on this already. We need to find the balance. How much does the network operator do on their own, vis-a-vis how much can we help them? It needs some business development.
organisation that would be serving not only one network, but several.
want to keep DOCSIS competitive, we support them by continuously introducing new technologies to enhance its capabilities. Then there is that midsection, who have long-term vision to overbuild with fibre, but who agree that it’ll take probably 15, 20 years. Those are the targets for outsourcing opportunities. Also, we have to keep in mind one more pretty important factor, the technology’s moving forward, but people are also moving forward and the level of competencies with DOCSIS and HFC RF transmission, that population is aging. Look at me. the knowledge base will be getting smaller, but that actually puts you in quite a good position again. HN: Operators are going to gradually reach a point where the critical mass of resources available is no longer adequate to run a quality operation. There aren’t enough competent people anymore. That sounds like there would need to be pooling of competencies that can then serve several larger network bases. I visualise another outsourcing activity. We could take over some of that engineering headcount, integrate them to a new
You’re tackling an uncomfortable subject that nobody really wants to confront. Where does AI fit into all of this? HN: With fibre becoming increasingly widespread, coax can no longer be considered the sole option for ultra-high- speed broadband. Which means that you have to compete in other ways - you have to differentiate on price, quality, network and service availability. Service is the main area; if you are failing on that one, you will be facing churn. And that’s when AI, machine learning, analytics and automation come into play. There are untapped opportunities to detect network anomalies before they impact your service, and this is another area where we have historically been the innovator. We have been talking about intelligent networks for 15 years. We need to continue to push the envelope and find better analytics to detect underlying issues in your infrastructure. With all this data that we can now collect from the network elements, cable modems, at home, the possibilities are limitless.
90
SEPTEMBER 2025 Volume 47 No.3
Made with FlippingBook - Online magazine maker