More
    HomeInsightsThe advent of AI and machine learning

    The advent of AI and machine learning

    -

    The increasing ubiquity of artificial intelligence (AI) and machine learning is extending beyond search engines and finance to telecoms, as operators look to new systems to improve their processes, potentially reinventing how networks are built and deployed in the process.

    David Eden, Future Technologist and Product Innovator at Tata Communications, says some of the most far-reaching uses for AI for the mobile industry are in network infrastructure design, maintenance and security. 

    Currently, there a limited number of people within the telco industry with the knowledge, expertise and experience required to design a network successfully, but Eden says that while these people are highly qualified, they are still human. Their decisions and designs are based on habit, personal preference and bias, all of which can come as a hindrance rather than a help. With machines, the same factors wouldn’t apply. 

    Eden says: “This means that by pouring the existing expertise in infrastructure design into AI algorithms, machines will discover new design processes and techniques which would never have occurred to humans.”

    This has the potential to increase efficiency and revolutionise the industry as a whole. However, Eden concedes that humans may not necessarily even understand the set-up and topology of these new AI-designed networks because of how sophisticated they may be. 

    This is just one area that operators and vendors see as being potentially transformed by AI and machine learning.

    “The case of machine learning and/or AI is very clear as another step to becoming a fully digital telco,” Phil Jordan, Global CIO of Telefónica, tells Mobile Europe.

    “Digitalisation for telcos is fundamentally about making the business model and customer experience real-time, automated and end-to-end across our business.”

    AI and machine learning are often used interchangeably, and while there is an overlap, there is a difference between the two. Machine learning is where a platform can be ‘taught’ using data, with its algorithm continually developing as it keeps receiving information. While AI uses data, and perhaps even machine learning, in order to act automatically.

    AI and machine learning are an extension of data analytics – an area that mobile operators have already been exploiting for a number of years. As Jane Zavalishina, CEO of Yandex Data Factory explains, operators have been using analytics for numerous benefits including churn reduction, marketing ROI and improved customer experience. However, she says current analytics methods being used by many are often underutilising the data at their disposal and not delivering the expected benefits. 

    She says: “Operators tend to aggregate data and examine the datasets to uncover useful business information – to support human-decision making. While providing certain business value, this approach is flawed.”

    Zavalishina suggests that humans using big data often generalise the information to a point where it can lose a lot of its power.

    “Human brains lack the processing power to interpret the sheer volume of data that operators must manage,” she claims.

    This is why machine learning, which can not only analyse data, but also predict, recommend and make automated decisions, is a step up from existing data analytics solutions. 

    She says: “It does not require the operator to have a deep understanding of the domain itself, nor to have data interpretation capabilities. The key is simply the provision of relevant historic data, which operators are blessed with a wealth of, for the ‘machine’ to interpret, test, act upon and then refine its outcomes.”

    Jordan says Telefónica manages massive data flow and has traditional ‘decisioning’ engines throughout the customer life-cycle, but says that AI and machine learning will lead to a day when the company makes ‘automated decisions’, whether this is in personalisation, customer offers, services and even in optimising the company’s own operations.

    Dealing with network traffic is another area in which machine learning could help mobile operators. “Moving to an automated model for traffic management will ensure customers enjoy seamless connectivity as capacity can be deployed where it is needed most,” says Joe Marsella, CTO EMEA at network strategy firm Ciena. 

    And on the network security side, Wandera, which works with the likes of BT, Three, Orange and Deutsche Telekom, uses machine learning to analyse billions of daily inputs across its network to detect zero-day mobile attacks and protect sensitive company data from leaking.

    It also uses AI to spot malware, leaks and suspicious user behaviour by training its algorithms to identify patterns and anomalies, and then automatically execute the response to that particular breach. 

    “This allows us to block over seven million threats a year for our customers, reducing the risk of security breaches on mobile devices,” says Eldar Tuvey, CEO of Wandera. 

    As humans are often thought of as a network’s greatest security weakness, Tata’s Eden believes that AI could be used to train human beings to make them more security conscious. 

    “AI can help us predict how people behave and identify the vulnerabilities that these behaviours are likely to create, improving our ability to predict and find the possible ways in,” he suggests.

    Meanwhile, machine learning could also help operators to predict customer behaviour so that telecoms companies are able to take proactive corrections actions.

    “For example, they would be able to detect when a customer is prone to [leave] the company and may propose actions to retain them,” says David Zakkam, relationship head at big data analytics firm Mu Sigma.

    As AI and machine learning technology becomes more sophisticated, the number of use cases for mobile operators is set to grow. Those companies who do adopt AI and machine learning may hold a crucial advantage over their competitors in the years to come.