Transforming telecoms through the adoption of artificial intelligence

Vishal Mathur, Global Head of Engagement at the Telecom Infra Project (TIP) and SONIC Labs Strategic Advisory Board member shares his thoughts on the latest trends and potential future developments surrounding artificial intelligence (AI) in the telecoms sector and the adoption of Open RAN technologies.

Throughout 2024, AI has been the deep tech generating the biggest buzz across the globe. Faced with an ever growing demand for fast and reliable connectivity, the telecoms sector is increasingly embracing AI innovation to unlock the potential benefits that this fast emerging technology promises. According to Fortune Business Insights, the global AI in telecommunication market size was valued at USD $2.36 billion in 2023 and is predicted to grow from USD $3.34 billion in 2024 to USD $58.74 billion by 2032; a compound annual growth rate (CAGR) of 43.1%.  

Enhancing the customer experience

Research from McKinsey shows that in 2024, 73% of senior telecom executives rate customer experience as a top priority. This is reflected in one of the biggest trends we’re seeing, with Generative AI increasingly being adopted by telecom operators to improve services to consumers and enterprises. This is being driven, particularly in North Asia and the Middle East, by significant investment in language specific Generative AI models and applications, as well as huge investments in cloud computing, data centres and server processing power. 

Within this context, telcos have a  huge opportunity to leverage AI innovations to transform contact centre operations, corporate communications and enterprise applications, enabling them to connect with customers in a tailored, personalised and rapidly responsive way – for consumers and business customers alike. Moreover, AI can help telcos to adapt and  generate new and meaningful services by providing the opportunity to develop new product strategies and introduce new revenue streams.

Addressing cost-effective consumption challenges

Additionally, many of the large telcos are looking to AI to help lower overall network operating expenditure and ensure future networks are responsive to real-world data consumption by developing software-based automation in their RAN (including Open RAN) and Open Transport networks. Open interfaces allow for increased choice in how telcos implement software-driven automation and how AI / machine learning (ML) could be implemented into RAN and Transport networks. 

Within their RAN networks, including Open RAN, mobile network operators are encouraging incumbent vendors, new entrant RAN Intelligent Controller (RIC) vendors and an ecosystem of AI/ML app developers, to develop non-real-time and real-time applications. These apps will directly manage and optimise the performance of telecom RAN networks, based on standards defined in the industry. Moreover, encouraged by the movement towards open and disaggregated networks, telcos are able to evolve from legacy network macros and proprietary Self Organising Network (SON) applications towards a mix and match approach of applications and RIC platforms that integrate across Open RAN interfaces into their radio unit-distributed unit-centralised unit (RU-DU-CU) architectures.

It is becoming increasingly clear that integrating AI into telecoms networks is a developing area where more work needs to be done; allowing Open RAN to become automated and cloud native, and used to drive forward a better way of managing Open RAN networks. What is encouraging is that we are already seeing multiple vendors developing proprietary and open RICs, as well as a growing ecosystem of AI/ML application developers.  

Rigorous testing and collaboration holds the key to successful adoption of AI technologies

Beyond ORAN Alliance’s standards development, the Telecom Infra Project (TIP) is facilitating the testing and commercial readiness of rAPPs/xAPPs {applications that run on a RIC platform to improve the operational efficiency of a radio access network) and RIC platforms, created by a global community of large and small vendors, and in response to a set of prioritised use case requirements defined by TIP members. 

However, the industry needs to work through various challenges before rAPPs/xAPPs can be commercially rolled out into a network. These challenges include taking a standardised approach to onboarding applications to RIC platforms, the ability to easily port applications across different RIC platforms, developing approaches to handle conflict between concurrently running applications, and undertaking standard carrier grade test plans that stress test the performance of applications under real world conditions.

Championing the development of data-enabled and open future networks is a key mission for Digital Catapult. Through its SONIC Labs programme, and in cooperation with other high capacity labs from around the world, Digital Catapult is working with TIP to provide the necessary test cases, test environments and toolkits for telcos and the vendor community to carry out Open RAN RIC and rAPPs/xAPPs testing, and to develop credible pathways to address adoption challenges.  

By focusing on this critical R&D activity, Digital Catapult is playing a major role in developing multi-vendor end-to-end networks to drive diversification in the UK telecoms market. With its unique, independent indoor and outdoor test environments, multiple Open RAN configurations including VIAVI and Keysight test toolkits, an ability to form digital twin environments, and a world class specialist technical team, SONIC Labs is well positioned to lead the UK and international industry efforts in carrier grade testing and industrialisation of rAPPs/xAPPs and RIC platforms.  

With UK Government support, Digital Catapult, TIP, UK telco operators and vendors, have the collective opportunity to drive this industry effort towards a path of establishing secure, cloud-native, automated, AI-driven open networks, and continue to support supplier diversification and a global marketplace of well tested, commercial AI applications. The work that SONIC Labs and TIP can do together can alleviate the time and effort that telcos will spend in their own labs to carry out this testing, and offer technical insights and blueprints that if accepted by telcos, can move the needle from pure R&D, to OTIC based integration testing, to actual commercial readiness of these Open RAN compliant solutions.  

We are already witnessing how AI can significantly contribute to the future success of the telecoms industry and help to address the challenges it faces, from improving customer experiences and network performance, streamlining operations and accelerating the adoption of technologies such as Open RAN – and I’m excited to see how we build on this and where our collective AI journey leads in the coming months and years.

Find out more

Visit our SONIC Labs section to find out more about Digital Catapult’s work with open networks. Additionally, if you are an investor looking for investment opportunities within the telecoms sector, find out more by applying to attend our Investor Seminar on Future Networks on 15 October.