Artificial intelligence (AI) is becoming an increasingly important technology for telcos to support their network operations.
According to a new report from Omdia, a sister company to Light Reading, AI tools have enabled telcos to manage the surge in customer engagements during the pandemic by providing self-service capabilities. AI has also improved network management.
"By 2022, CSPs will accelerate their use of AI," the report said.
"Though still in the early days of its adoption, AI is driving positive outcomes as CSPs' confidence levels in the technology's ability to drive value to operations has increased in the last 12 months. This trend will fuel increased demand for AI to drive more use cases, including network operations, cybersecurity, and sales as top use cases."
However, Omdia advises telcos to look beyond the initial success achieved with early telco AI use cases.
"To tap into the value that AI can bring, CSPs should begin to invest in effective ways to scale its use and implementation," the research company said.
"This is a key focus for CSPs as they prioritize investment in capabilities including low- and nocode, to speed up development, and cloud-native serverless, to affordably scale deployment of AI."
Yet while scaling AI implementation is important, Omdia warns that this must be balanced with governance and transparency "to ensure a well-regulated and ethical use of the technology."
In its report, "2022 Trends to Watch: Telco AI," Omdia notes that while network AI use cases continue to dominate, momentum around other use cases is increasing.
It also found that investment in low- and no-code frameworks is a top priority for AI development, and that a cloud-native serverless environment on public cloud will be the preferred environment to deploy AI workloads.
Omdia's recommendations for telcos are that commitment from leadership will be key to long-term success.
"Scaling AI operations, including development and deployment, will benefit from having C-level representation on the board to drive budgetary needs for AI projects," the research company said.
It also advises them to develop and follow a standard AI development lifecycle from "ideation" to production.
"In place of creating AI workflows from scratch for every new use case, having a standardized workflow that can be automated will be beneficial when it comes to scaling AI solutions," Omdia said.
Other suggestions include investing in an AI development platform that supports AI at scale and establishing strong data architectural and management practices.
"Data management and governance must become core to CSPs' operational practises," Omdia said. "They need to build a robust data strategy that seeks to understand what data assets are available and where they are located, and that provides access to the data."
- Telenor does AI deal with Google despite its big-tech fears
- Infovista plays to the crowd with AI-based 5G RAN planning
- NTT scrubs 'dirty networks' with AI threat sensor
- Juniper Networks extends AI-driven support to entire Junos portfolio
— Anne Morris, contributing editor, special to Light Reading