Skip to main content

Stop! Are you doing it wrong? There’s a better way to apply AI to telecoms

DevOps has been hailed as a transformative technology for Communications Service Providers (CSPs). Long used in the IT world, DevOps has crossed into telco, enabling CSPs to dramatically reduce the time to develop and implement enhancements to systems and processes.

Yet a new Analysys Mason report* claims DevOps processes are still overly reliant on manual intervention, with people configuring current systems or coding to support new functions. This means systems and processes are not kept up to date and specific use cases are not addressed precisely.

The solution is to apply Artificial Intelligence (AI) to automate development processes by dynamically detecting and reacting to changes without manual intervention. But even this is not enough, says the report. The author points out that often, end-to-end processes are not being optimised, only the machine learning systems themselves are being enhanced.

Three requirements for CSPs to watch for

Using AI to enhance their business and operational processes is an important development for many CSPs looking for the most cost-effective ways to evolve their networks to offer competitive ultra-fast broadband. For these CSPs, AI is an attractive proposition to help position themselves as a leader with innovative digital services that are high quality and competitively priced.

But how is AI technology best applied?

The Analysys Mason report makes three recommendations to help CSPs choose the right approach:

  1. Ensure AI integrates tightly with the vendor’s current applications and can support professional services using cloud-based tools. This is vital to speed up CSP deployments and ensure a service-led approach
  2. Choose a single AI ecosystem that ensures systems can be automated fully, end-to-end, to create processes that are more efficient, faster and lower cost
  3. When adopting an AI ecosystem, CSPs are advised to provide their own, independent support. Creating in-house teams trained and able to use the ecosystem, and using open-source tools will avoid vendor lock-in.

The report also says that CSPs will benefit if there is a library of use cases programmed and configured to ensure automated updates are aligned to the required business outcomes. Each use case needs to target a specific business issue.

Analysys Mason says Nokia AVA does it the right way

Helpfully, the Analysys Mason report also includes an analysis of several solutions available on the market. Its investigation reveals that: “Nokia AVA is the most mature offering in the TAE (TAE = Telecoms AI Ecosystem) market”.

Nokia AVA integrates Automation, Virtualization and Analytics to deploy advanced services to CSPs. As a cloud-based platform, AVA offers centralized collection and processing of data from across the network and makes it available to any processing or reporting tool. This enables CSPs to achieve near-perfect network availability while providing ultra-low latency and high throughput.

The solution meets the three requirements outlined in the report – it provides AI-based support with automated end-to-end analytics to help CSPs make smart decision on how to evolve and optimize their networks. It is open-source based, enabling CSPs to integrate their own or third-party tools and solutions via open Application Programming Interfaces (APIs). A wide range of commercial models is also available, enabling CSPs to adopt Nokia AVA as their single AI-based solution without needing any upfront CAPEX investment.

The Analysys Mason report also points out that: “AVA lends itself to operators who want to apply AI and analytics-based use cases”. The solution offers more than 200 use cases available from a library based on real CSP needs. These use cases give CSPs risk-free ways to gain the benefits of AI technology. For example, machine learning can provide very detailed sub-cell insights and make automatic optimization recommendations to boost network performance.  This approach helped Hutchison 3 Indonesia to increase spectral efficiency by 17 percent.

For CSPs looking to evolve their network to deliver, competitive, leading digital services, with minimal risk and low costs, Nokia AVA offers the right approach. At least that’s what Analysys Mason suggests.


AI ecosystems enable CSPs to automate their development processes and achieve near real-time deployments – schematic from Analysys Mason report “Telecoms AI ecosystems in digital transformation: increasing the level of automation in critical processes”*

Share your thoughts on this topic by joining the Twitter discussion with @nokia and @nokianetworks using #AI


* Telecoms AI ecosystems in digital transformation: increasing the level of automation in critical processes, Analysys Mason, July 2018

Andrew Burrell

About Andrew Burrell

Andy is responsible for Nokia’s Digital Operations’ marketing.  After more than 25 years in the telecoms industry he remains fascinated by the possibilities of technology, in particular the potential of 5G and Artificial Intelligence to transform networks, operations and business.  He loves to use his various devices to keep up with social media, news, and above all, the football results. Finally realizing that the phone call from Arsenal FC was never going to come, he has given up playing and now prefers to watch his kids chase their own dreams from the side of the pitch.  Tweet me @andyburrell or connect via

Article tags