Operators can gain energy efficiencies by using AI technology, but have to holistically consider AI and AIOps
While the focus of the Telco Sustainability Forum (available on demand here) is obviously around the technology and business strategies operators can use to drive energy efficiency, this is part of a bigger whole. Many emerging tools that help operators reduce power consumption and realize system-level energy efficiencies hinge on the use of artificial intelligence (AI) and closed-loop automation; and any strategy that involves the adoption of AI in pursuit of closed-loop automation should include sustainability targets, it should also be comprehensive and mindful of that bigger picture.
Nokia’s Nabil Lahyani Hamidi, head of analytics delivery services and care in the Cloud and Network Services unit, emphasized the importance of an aligned AI technology and operationalization strategy (AIOps) in a conversation with RCR Wireless News. “It’s very easy to claim that we have AI, we have [machine learning], bringing it into the picture, but actually most operators or customers who I’m working with…the most tricky aspect is the organizational/cultural shift because that requires a change in terms of processes, in terms of tools and in terms of also the capabilities we have in our organization.”
He described three waves that operators will have to navigate in pursuit of network automation:
- The adoption of business intelligence tools that use data analytics to feed into dashboards and reporting and result in a new type of insight—”You cannot be blind.”
- AI models drawing from data lakes enabling different use cases, albeit use cases that perhaps exist in silos.
- And, third, shifting from using data insight to facilitate more efficient manual operations to automating those operations.
“All that requires this AIOps concept in order to operationalize it,” Hamidi said. And, he added, there needs to be an unwavering focus on tying these technological and operational changes to clear business objectives.
Specific to the energy efficiency and sustainability angles, Hamidi noted that energy consumption is between 15% and 30% of operator opex in mature markets, and up to 50% of opex in developing markets. With the proliferation of 5G—which is more energy efficient than 4G on a one-to-one basis—”the reality is different. Due to the high degree of density of 5G networks, without any intervention, it will be really hard to act and bring energy efficiency. It will require an end-to-end approach starting by bringing an AI-driven network design and solutions. And, afterwards, AI-driven equipment control.”
Hamidi also made the point that investments in sustainability do not equate to compromising traditional network performance KPIs. In fact, he said, “It’s one of our differentiators because we do care about network quality. We do care about how people are experiencing or using their services.”
In terms of takeaways, Hamidi emphasized operators need an end-to-end strategy, and they need to understand that AI technology has to be implemented along with an AIOps structure. “We need to have a clear strategy that is aligning different departments, then how to operationalize it, and how to learn from that experience. Such complexity from technology and the fast pace requires transformation.”
