As Artificial Intelligence (AI) continues to revolutionize industries worldwide, the telecom sector is no exception to this transformative trend. The big question we need to ask ourselves is how we leverage this technology in telecom for substantial gains in productivity.
While the internet generates a staggering amount of data each day, it is important to recognize that a significant volume of network data is also generated in the backend. This valuable data can be leveraged for various purposes, such as network optimization, predictive maintenance, fraud detection and enhancing customer service.
With network differentiation progressing towards prioritizing user-level quality of experience, softwarization and cloudification, mobile operators are presented with valuable new opportunities to improve customer experience. However, to do this, they must first embrace a fundamental shift in their operating model.
In today's telecommunications landscape, there is a need for harnessing the power of AI and automation to deliver speed of execution, operational efficiency, cost-effectiveness, and ability to innovate at parity or better than the competition.
Rakuten Symphony has been embracing this change since its inception and enabling operators to centralize configuration management, data collection and visualization, automate anomaly detection, fast root cause analysis (RCA) and next best action (NBA), and generate meaningful insights and actions for closed-loop automation.
The result: Our obsession with automation has created our biggest differentiator: speed, enabling us to deliver results in minutes instead of days or months.
Rakuten Symphony’s Symops Service Assurance, a key component of Symworld™, harnesses the power of AI and automation to optimize every stage of the service provider journey. It consists of revolutionary apps that work to automate operations and incident resolution.
The two fundamental system designs we had to put in place were establishing a vendor-agnostic unified data model based on open standards and prioritizing real-time data availability to enable complete autonomy and seamless automation.
Let us take a closer look at applications contained within Symops Service Assurance that are delivering the much-needed transformation:
The journey of automation and AI begins with implementing a cloud-native observability framework. With Rakuten Symphony Symworld™, everything seamlessly integrates into a unified and customized micro-frontend/microservices-based dashboard portal. This portal empowers operators to efficiently monitor and manage the network from a single, comprehensive view.
The observability layer enables operators to ingest and process valuable data streams from various network domains, including transport, security, core, RAN, and elements spanning multiple technologies. These data streams are then normalized to extract fault and performance metrics, allowing the generation of key performance indicators (KPIs) and incidents. This data is then passed through an RCA (Root Cause Analysis) engine, such as the Watchtower tool, which empowers users to establish RCA rules and policies for detecting and initiating necessary actions. The Symworld™ platform acts as a foundational layer for aggregating data from all sources. This empowers operators to proactively take actions and ensure service availability by leveraging real-time insights.
From Day 1 configuration to Day 2 provisioning management (e.g., when a cell or core network element is deployed), the Configuration Manager monitors network configurations against predefined parameters and generates compliance reports.
It also enables automated configuration provisioning enabling Zero Touch Provisioning (ZTP) to be seamlessly achieved through northbound integration with Symworld™ Orchestrator.
Incidents are generated based on alarm correlation and performance threshold breach, enabling users to take proactive measures. For non-compliant parameters, they are displayed on the dashboard, and users can access audit details to review the list of non-compliant parameter details. This ensures a proactive network configuration and operation approach, enhancing overall efficiency and performance.
Fault Monitor offers fault/incident management dashboards for operators to monitor and address network issues from the NOC (Network Operations Center). Based on the occurrence of faults/incidents, the platform generates meaningful incidents and provides tools for effective incident management. Operators have the ability to integrate the incident management system with a ticketing system for efficient trouble ticket creation and management, or leverage Watch Tower, an auto RCA (Root Cause Analysis) and policy tool. Operators can also utilize Automation Studio to execute scripts and automatically remediate incidents as soon as the auto trouble ticket is created, enhancing operational efficiency and response time.
Operators can define and implement multiple RCA policies, allowing incidents or network faults to trigger these policies. Triggers can also be based on performance KPIs. By utilizing these triggers, operators can navigate through the RCA trees and initiate necessary actions, such as generating a trouble ticket through the ticketing tool, adjusting system configurations to improve KPIs, or integrating custom actions with Symworld™ for automation purposes. This simplified interface facilitates the achievement of automation objectives for operators.
In terms of Performance Monitoring, RAN and core network KPIs are closely monitored at regular intervals. These metrics are then visualized through dashboards or integrated with a northbound system. Symworld™ also brings the option of using its BI solution for driving intelligence and dashboards, but it could also be linked with the operator's BI system.
Operators can set threshold conditions within the performance management (PM) capabilities to generate alerts on KPI degradation and automatically create incidents with auto trouble ticket rules configured for immediate attention as soon as an incident occurs.
The Performance Monitor monitors violations and can be utilized by the auto RCA engine for closed-loop automation. The KPI builder allows operators to customize KPI formulas based on integrated vendors' metrics and desired quality measurement approaches. Anomaly detection and KPI forecasting also leverage AI-assisted observation through ML (Machine Learning) models.
As part of our service assurance portfolio, we have a service desk incident management engine that holds extensive data on root cause analysis (RCA) for incidents related to change management and method of procedures. Our service desk system is specifically designed for telco operations and provides AI-powered insights and troubleshooting guidance to network operations center (NOC) engineers. This valuable data can serve as a training ground for future Telco AI foundation models.
Our goal at Rakuten Symphony is to create a robust ecosystem that aims at shifting the focus from investing billions of dollars in network upgrades to investing intelligently in the drivers of productivity.
When it comes to an AI-powered Telco, AI is simply the next software algorithm to achieve operational efficiency. The ultimate objective is to reach the next level of automation maturity, leveraging AI algorithms to analyze large, unified data sets for success.
"At Rakuten, the number one problem we solved was speed of execution. And the key to achieving speed lies in automation. Automation only works if you have a unified data understanding. Once you have a unified data understanding, you are well positioned to apply AI models to that data."
For more information on our transformational journey and an in-depth breakdown, download our whitepaper here on Mobile as a Software™.
