It is no secret that artificial intelligence (AI) is a technical marketing whitewash. Many companies claim that its algorithms and data scientists enable a differentiated approach in the networking infrastructure space. However, what are the practical applications of AI for connectivity and, in particular, 5G? From my perspective, it encapsulates four key areas. Here I will provide my insights into each and highlight what I believe is the practical functionality for operators, subscribers and equipment providers.
Automation is all about reducing human error and improving network performance and uptime through activities such as low to no-touch device configuration, provisioning, orchestration, monitoring, assurance and reactive issue resolution. AI promises to deliver the “smarts” in analyzing the tasks above, steering networking to a more closed-loop process. Pairing all of this with 5G should help mobile service providers offer simpler activations, higher performance and the rapid deployment of new services. The result should be higher average revenue per subscriber (ARPU) for operators and a more reliable connection, and better user experience.
Over time, I believe AI will evolve to enable network operators to move from reactive to proactive issue resolution. They will be able to evaluate large volumes of data for anomalies and make course corrections before issues arise. 5G should enable networks to better handle these predictive functions’ complexity and support significantly more connected devices. We’re beginning to see AI-powered predictive remediation applied to the enterprise networking sector to positive results, via some tier one carriers and 5G infrastructure providers such as Ericsson. In my opinion, one of the most significant impacts of AI in mobile networks will be the reduction of subscriber churn. That is a huge consideration—carriers are spending billions of dollars building fixed and mobile 5G networks. They must be able to add and retain customers.
Digital transformation acceleration
One of the pandemic’s silver linings is the acceleration, out of necessity, of businesses’ digital transformation. The distributed nature of work from home has put tremendous pressure on corporate and mobile networks from a scalability, reliability and security perspective. Many connectivity infrastructure providers are embracing AIOps for its potential to supercharge DevOps and SecOps. AI will also help operators better manage the lifecycle of 5G deployments from a planning, deployment, ongoing operations and maintenance perspective. For example, China Unicom leveraged AI to transform how it internally manages operations and how it interfaces with partners and customers. In 2019, the operator reported a 30% reduction in time to product delivery and a 60% increase in productivity for leased line activations.
Though AI is hyped all around, there is particular synergy with 5G. Mobile networks are no longer just a “dumb pipe” for data access. AI can improve new device provisioning, deliver high application and connectivity performance, accelerate digital transformation and provide exceptional user experiences. For service providers, I also believe AI and 5G will result in operational expense savings and drive incremental investment in new service delivery. In my mind, that is a win-win for subscribers, operators, and infrastructure providers alike.