• Major German telco O2 Telefónica has been testing an AI-based network planning tool
• The test results were so good it is now using the tool to plan its live network upgrades and expansions
• The new system delivers “twice as many accurate predictions” as previous network planning tools, according to O2 Telefónica

Major German telco O2 Telefónica (aka Telefónica Deutschland) is using an AI tool developed with Google Cloud technology to predict mobile data usage patterns that will enable “more precise capacity planning, targeted investments and optimised network expansion for a better customer experience,” according to the operator. 

The AI system, which was tested last year, generates “twice as many accurate predictions” than previous planning tools and has an accuracy rate of more than 90%, according to O2 Telefónica. 

The operator’s network planning team uses the AI tool to analyse data traffic statistics – current and historical – collected from its 28,000 radio access network sites and factors in other information, such as mobile package tariff changes. The AI analyses about 250 Gbytes of data every few minutes and is able to provide a forecast of expected data traffic patterns for the coming months and years in just half a day: Conventional network planning systems take weeks to perform such calculations and with less accuracy.  

“We are setting new standards in network expansion and network planning,” noted chief technology and enterprise officer Mallik Rao. “Artificial intelligence helps us to gain sound insights and make accurate predictions. In this way, we are expanding our network capacities in a targeted manner. We are expanding where our customers receive real added value. This means they experience a first-class internet and telephone experience on the O2 Telefónica network,” added Rao, who took on his expanded role in November last year. 

Mobile operators have never needed such accurate planning tools more. O2 Telefónica, which has more than 45 million mobile customers, says more than 5 billion Gbytes of data flowed across its network last year – five times more than in 2019 – and that it undertook 7,500 capacity expansion measures last year. According to the operator, the AI system enables it to focus its mobile network investments better than ever before. 

Network design efficiency and improved network planning were cited as some of the most important ways in which mobile network operators (MNOs) can benefit from the use of AI tools by the MNO respondents to TelecomTV’s industry survey conducted during the second half of 2024 – you can read the full results of that survey in our free-to-download DSP Leaders report, AI’s Impact on the Radio Access Network.     

- Ray Le Maistre, Editorial Director, TelecomTV