Vendors and communication service providers (CSP) have been discussing the benefits of artificial intelligence (AI) for some time now, but so far it has amounted to little more than talking points. However, with 5G networks coming online around the globe, advanced use cases being introduced and the profound impact COVID-19 has had on CSP operations, the need for AI is becoming ever more apparent. Expectations for proactive customer service and care also placed increased pressure on business support systems (BSS) and frontline employees. The complexity and real-time nature of network slicing, advanced digital services, and industrial IoT deployments place additional operational demands on CSPs that are best automated and managed by AI.
Although, it is early days for AI in CSPs, there has been activity across the board. According to a December 2020 TM Forum survey, 60-80% of CSPs are already incorporating, developing proofs of concept, or road mapping AI into customer experience, revenue management, service creation and management, and network planning and management. And while this is impressive, it is only the beginning. Nearly 60% of surveyed CSPs intend to increase AI investments in 2021. Here are a few areas ripe for AI adoption.
Chatbots and automated responses are not new concepts. Nevertheless, we should not ignore the profound impact AI-enhanced customer service tools can have on customer experience, helping to better serve the business interests of the CSP. Anticipating customer problems, proactively making better customer offers, and delivering a seamless experience across the entire customer journey are all areas that AI can enhance.
AI-enabled customer experience solutions can increase revenue opportunities and limit loss through personalized offers, lead generation and churn management. CSPs have a vast amount of data at their disposal, and the ability to analyze these deep troves of external and internal data quickly, efficiently and in real-time will help CSPs identify new potential customers with personalized offers. For example, charging engines could use AI to dynamically change rates for customers based on behavior, location, service or any variety of other parameters available. Similarly, AI will be able to help CSPs provide personalized offers in response to customer service issues. By monitoring customer sentiment through service interactions, AI can pick up on cues that may indicate propensity to churn. The sooner the CSP can pick up on this, the better its ability to mitigate, stemming customer loss and enhancing the customer lifetime value.
AI can enhance network operations on existing networks through the optimization of network rollout, network assets deployment, field force, as well as predictive maintenance. China Mobile recently outlined many of the results it has been able to extract from AI-enhanced network operations. These include increases in outdoor downlink and indoor networks speeds, as well as automating many of the fault alarm scenarios. While these represent obvious impressive operational improvements, the real impetus will come with 5G networks.
As 5G evolves into its more advanced stages with network slicing and other complex use cases, the need for AI will become increasingly important for the successful management of these services and use cases. Being able to instantiate, operate, and retire network slices and advanced services automatically, quickly and without error will be critical to the success of money-making 5G services. The number of devices and required resiliency of more advanced 5G use cases will require levels of automation and predictive action previously not needed. Traditional operational models and resources are insufficient to provide this level of service. AI will be a necessary tool for CSP operations to minimize risk of fault or error. CSPs will need to begin building out this capability in advance of these services. They will need to operationalize these capabilities on existing networks and services to ensure resiliency when they go live with new services. Eighty percent of CSPs view using AI to automate network operations as a top priority.
The future is in the cloud. This is true for CSPs as much as it is for other enterprises. And hand-in-hand with network operations, the increased adoption of cloud technologies will push CSPs towards increased adoption of AI in their operations. As cloud-native deployments, the use of cloud infrastructure and an intelligent edge proliferate, orchestration of applications and operational functions will be more complicated and require more automation. AI will be critical to successfully managing all this complexity efficiently and optimally.
CSP applications, such as charging and inventory control, will need to be deployed in a distributed, cloud-native fashion to ensure that service levels in latency and resiliency are maintained. Managing these distributed applications will be practically impossible without automation and AI. As closed-loop network automation takes hold, CSP attention will naturally shift to edge-deployed AI for optimized service delivery.
Laying the foundation for long-term success
As is evident, there are many use cases that warrant AI consideration and investment by CSPs. The forthcoming growth in advanced 5G use cases show us that AI is a necessary part of the CSP operational model going forward. A June 2020 Boston Consulting Group study predicts that infusing AI throughout the CSP value chain can increase revenues by up to 10%, while reducing costs by up to 20%. That is a significant positive impact financially and operationally. Other operational benefits include streamlining processes, optimizing operations, and ultimately, freeing up resources to spend time on higher value issues as well as innovation investments.
COVID-19 has revealed how critical automation is for key functions and processes, even without 5G on the horizon. As the foundational technology for automation, AI will enable CSPs to make important operational decisions when human intervention is compromised or impeded. Even in today’s uncertain environment, it is safe to expect this year to be the year when we see AI become a more central part of CSP operations. CSPs need to be prepared.