With mobile network operators (MNOs) increasing their investment in 5G standalone (SA) networks, is mobile telephony on the verge of an evolution? There’s a strong argument that it is. 5G SA introduces a new network architecture capable of delivering many new capabilities that were not feasible with previous mobile generations, says Susan White, head of strategy and portfolio marketing, Netcracker Technology.
These include ultra-low latency, extremely high data rates, massive IoT connectivity, and network slicing that can be transformational for many industries. It is a connectivity revolution. And it could be extremely lucrative. According to a McKinsey discussion paper on mobility, healthcare, manufacturing, and retail, 5G connectivity could boost global GDP by $2 trillion (€1.87 trillion) by 2030.
The report’s authors explain, “Connectivity will enable businesses to do more. Enhanced broadband will make streaming, downloads, and data exchange lightning fast. Because they require less power, LPWANs will extend the battery life of the devices and sensors they connect, making it viable for the Internet of Things (IoT) to scale up like never before. Ultra-low latency and strong security will create the confidence to run mission-critical applications that demand absolute reliability and responsiveness even in vital infrastructure systems and in matters of life and death.”
This is an unmissable opportunity for MNOs. But along with the benefits come major challenges. The quantity of data flowing across 5G SA networks will be exponentially greater than current volumes. And much of this data will not even ‘belong’ to the MNO.
Many MNOs are already embracing analytics, artificial intelligence/machine learning (AI/ML) and automation to some degree to prepare for the data-driven future. This trio of factors is occasionally called A3, and according to analyst STL, it can help telcos address six problems:
- Making sense of complex data – identifying patterns, diagnosing problems, and predicting resolutions.
- Automating processes, orchestration, and completing tasks.
- Personalising customer interactions – analytics and ML can be used to understand customer data, create segmentation, identify triggers, and prescribe actions.
- Supporting business planning – forecasting demand and optimising use of existing assets.
- Augmenting human capabilities with natural language processing (NLP), text analytics, etc., allowing bots to help customers and employees achieve tasks more quickly and efficiently.
- Identifying frontier AI solutions – AI solutions which have specialist uses within a telco but are not yet widely adopted.
The problem is how to harness, manage, process, and act on massive amounts of data at scale. When done right, it will revolutionise the telco business. Analytics and AI will help CSPs bring order to the chaos resulting from the tsunami of data. However, without the right data, the levels of automation CSPs need to achieve will simply not be feasible.
Today, when an MNO (or MNO customer) wants to make meaningful data-driven decisions for well-defined business use cases, they need to extract the relevant data and ensure the data is accessible and in the correct format for the analytics use cases.
However, the data needs to be extracted from many sources BSS, OSS, ERP and other systems. This can be a challenge: each system typically uses a proprietary data model that is incompatible with the others, and is often unsuitable for analytics use cases. It makes this process extremely difficult.
Unlocking this data and transforming it into useful data is at the heart of a data-driven strategy with the following key principles:
Identify value-driven use cases
AI strategies need to start by defining use cases that solve the telco’s biggest issues with measurable KPIs. These might include increasing customers’ lifetime value, reducing churn, improving chatbot performance with Large Language Models, generating new revenue streams by providing more personalised marketing, preventing fraud, and more.
Unify the data into a meaningful structure
Extracting and preparing data from many different IT systems into a unified and usable format is critical to a data-driven business. This is accomplished by using an analytics data platform with a model that sources the right data across the entire multi-vendor telco data and analytics value chain, and then transforms the data to make it easily consumable by analytics tools. The analytics data platform must easily integrate with the existing MSO infrastructure including data lake, data warehouse and online data storage.
Integrate with large language models (LLM)
The integration of LLM (e.g ChatGPT) into the telco data environment will transform the chatbot experience. It should enable better support for more complex problems. For example, LLMs will help agents to deal with multiple languages and provide better context. In fact their use will impact the entire business to drive operational and business efficiency.
Make it easy for business entities to use the insights
Once insights are available, it can be a daunting task for business users to access them to create reports and dashboards. This mode of operation has to change. This can be done by providing intuitive self-service analytics with out-of-the-box data marts and visualisations tailored for specific users. These tools will give any user access to data that can improve business functions and services. Needless to say, these tools should be presented in a business-friendly user interface/no-code approach.
Build a unified AI/ML framework
With ready-to-use ML models and a blueprint approach, telcos can expedite the creation of ML-driven use cases and model retraining. They can use MLOps to streamline the process of deployment and maintenance of these models in production reliably and efficiently.
How might the above benefits translate into business growth? They can completely redefine every aspect of the customer experience retaining more customers, upselling relevant new services, and fixing problems before they are known. These improvements will help MNOs to be on a par with the experience people have come to expect from tech companies. And they can optimise the performance of highly complex networks and services highlighting coverage gaps, increasing operational efficiency, maintaining KPIs, and meeting sustainability goals.
Within the IoT space, there are important applications CSPs should consider such as ‘anomaly detection and predictive maintenance’. MNOs will be able to give enterprise customers access to AI-enabled dashboards so they can monitor the health of their devices in real time, and act on anomalies. Similarly, the system will be able to detect threats by scanning for unclassified devices or behaviors that might indicate an attack on the network.
The mobile industry is currently poised for a huge influx of data traffic, and much of it will be generated by machines and sensors, rather than people. In fact, IDC believes there will be 55.7 billion connected IoT devices by 2025, and that these devices will generate almost 80 billion zettabytes of data. To maximise commercial gains and thwart security threats, telcos and their enterprise customers will need a way to control the massive amounts of data the future will bring.
While 5G SA networks will unleash opportunities for MNOs, they will also result in a tsunami of data that will need to be harnessed, managed, and processed. Addressing this challenge and gaining the ability to act on the data requires CSPs to embrace automation, analytics, and AI/ML. It is only through ‘the three As’ that telcos will be able to fast-track business growth and profit in the 5G SA era.
The author is Susan White, head of strategy and portfolio marketing, Netcracker Technology.
Comment on this article below or via Twitter: @IoTNow_OR @jcIoTnow
This UrIoTNews article is syndicated fromIoT-Now