AI by Design – AI solutions solving the right problems

Our artificial intelligence is applied to solving the right challenges for telecom service providers, creating value where it matters the most. It is not about gimmicks or generic experiments. It’s not a general-purpose tool intended for any use case imaginable. Our AI solutions are embedded throughout the network, built by people with extensive AI and telecom expertise, and with an AI-first mindset and technology in every product or service. Our AI is deliberate and purposeful. We call it AI by Design.

Our AI solutions in action

AI-powered Network Services

Using predictive analytics, AI techniques and automated machine learning to deliver Ericsson’s latest Network Services. It’s the key to secure always-on networks in the 5G era.

5G-aware traffic management

With AI embedded in the RAN Compute software, 5G capable devices can be directed to the best 5G capable cells, keeping users with 5G phones in 5G coverage. Cells are ranked based on AI models and dual connectivity capability.

Up to 25% better 5G coverage

Augmented MIMO sleep

In Ericsson Radio System, AI algorithms run on the baseband to predict traffic patterns and autonomously turn off antennas as required to reduce energy usage. It’s also possible to combine with Cell sleep and the Low Energy Scheduler solutions for even better savings.

14% - average saving for a proof-of-concept cluster

Self-organizing networks

Intent driven proactive and reactive optimization workflows combined with AI to optimize Cell Clusters at scale. Machine Learning-based unsupervised clustering techniques is applied to discover Network Topology.

18% better network load redistribution

Cognitive performance diagnostic

Innovative, efficient and accurate AI-powered approach to detect daily issues affecting network performance.

15 min vs 1 week taken to automatically classify 100 000 cells with 2 weeks of data



RAN KPI degradation prediction

A solution that analyzes data coming from OSS. Using AI and Machine Learning, models can predict the possible throughput degradations or increased latency before they occur and automatically trigger corrective actions.

70% of incidents automatically prevented

IoT connectivity

Machine Learning assisted paging

Machine learning improves the existing patented MME features of adaptive paging so that it becomes topology-aware and reduces signaling, which in turn frees up capacity and defers investment for carriers.

80% reduced signaling (paging)

AI solutions that accelerate the transformation of service providers along three dimensions

Service providers are on a journey from legacy telecom service providers to becoming digital service providers. Ericsson’s overarching strategy is to enable this transformation in three principal dimensions – Relentless efficiency, End-customer experience, and New revenue streams. AI adds functionality and capabilities in all dimensions.

Drone inspecting oil rig

Relentless efficiency

Transforming and simplifying both IT and network environments and adding large amounts of automation to operations.

Communication services

End-customer experience

Customer insight-driven operations providing experiences that meet or exceed digital natives’ expectations.

home group of young people

New revenue streams

Delivering new services to new target groups, enabled by new network capabilities.

Network Service Intelligence at your fingertips - webinar

Are you planning to adopt advanced analytics in your operations? Are you ready for 5G? Here’s a chance to find out how we’re using predictive analytics, AI techniques and automated machine learning to deliver Ericsson’s latest Network Services. This is not just about continuous support: it’s the key to secure always-on networks in the 5G era.

On 12th March 2020 at 2PM

Join the webinar

AI solutions solving the right problems for service providers

Our AI solutions target service providers to address their challenges to maximize efficiency and end-customer experiences and create new revenue streams. To do this, our work with AI solutions is focused on the problems that we think really matter to our customers, rather than any problem. We have identified six main problems that we solve.

The world becomes increasingly dependent on superior network performance, both from a consumer and a business perspective. It’s the main driver of consumer satisfaction, and absolutely necessary when deeply integrated into industrial production processes. It needs to be always up, always on, always performing. We use AI in our networks to secure top-class performance.

Reducing network energy consumption benefits everyone – the consumer, the environment and the bottom line. We use AI to plan, manage and run networks more efficiently from an energy perspective.

Operations are key for service providers to realize business opportunities. Service providers need to create relentless efficiency in operations by becoming data-driven and predictive. We use AI to increase efficiency in operations while transforming to deliver on business KPIs rather than network KPIs.

Virtualization and cloudification promise improved agility, lower cost and is the foundation for new service creation. A cloud-powered network with a virtualized core is a competent tool for service provider transformation, but it also comes with challenges. We realize the value of the virtualization of core and cloud technologies with AI. 

Future customer experience is based on more than coverage and capacity. With diverse use cases, the definition of customer experience will be equally diverse. To meet these requirements, manual activities are not enough, and AI will enable service providers to create and assure experience no matter what those services are.

AI is not only about 5G and the future. It’s a powerful tool to bring out most of the infrastructure already in place, making sure that you can keep the cost down and customer experience high.

We have AI capabilities that solve major problems for service providers

We are not building a general-purpose artificial intelligence platform, trying to solve problems in every industry. Instead, we have developed our skills and capabilities in telco AI to address the challenges of service providers. Working closely with our customers’ challenges, data and products, we are providing them with a different approach to getting value from their networks. We have identified and prioritized four AI capabilities that make our AI as effective as possible.

Networks intelligence and automation, Research

Telecom & AI expertise

Our AI is anything but artificial, it’s built on a strong foundation of combining deep telecom domain expertise and data science- and AI knowledge. These skills always work hand in hand, close to customer problems, collaborating to solve them. This means that we know the right problems to solve and how to solve them with the right AI solutions.

Machine intelligence

AI-ready technology

Ericsson’s technology is AI-ready. We have made architectural decisions that make AI easy to deploy and use and that works together holistically across the network. Combined with our telecom domain expertise, this means that we can use AI where it makes sense, solving the right problems for our customers.


AI use case co-creation

We have learned through experience that co-creation in developing solutions together is critical. We develop our AI in the field together with our customers, using real network data, our domain expertise. Together with a portfolio of ready-made use cases and pre-trained algorithms, we adapt to various customer contexts and problems. In this way, we can solve the right problems for our customers instead of providing an untrained general platform or solutions to only address very specific challenges.


AI-powered portfolio

We are not a niche player, focusing on one use case. Instead, we build AI into every part of our portfolio where it makes sense, from baseband to centralized clouds. With AI spread out through all our products, we can see the whole end-to-end context and how we make every part of the network work together.

Dense Urban Street Macro

Unique access to data

Through access to multiple and diverse data sets from CSPs that we collaborate with, we can train our algorithms. This valuable data creates a unique foundation to build powerful algorithms, benefitting CSPs and their end-users.

AI to boost 5G implementation

5G is introducing new challenges for service providers. We have adopted AI across our product and service portfolio to make the 5G switch easy for our customers.

AI enabling always-on network

Achieving always-on network with our AI-powered preemptive support, addressing any challenge that might arise in good time and a personalized digital interface share AI-driven insights and information on new technology.

Find the report here

AI in 5G networks

AI stemming from the latest mobile communications innovations is the subject of our latest market research.

Read the highlights here

AI and machine learning in next-generation systems

Based on insights about future 5G systems and developments in manufacturing and ITS automation, this white paper reflects on the technical challenges that need to be addressed to fully capitalize on the potential of AI and Machine Learning.

Read the whitepaper

Telecom Analytics

It’s critical to deliver an outstanding customer experience despite increasingly complex operations. Artificial intelligence and Machine Learning help to manage the complexity and resolve problems proactively. Take a look at how Ericsson Expert Analytics uses Machine Learning models to search and resolve anomalies in the network so that the customer experience remains optimal.

Telecom analytics

The Voice of AI

Working in telecoms? Interested in AI, but not really that deep into what it is? This is the podcast for you!

Listen to podcast