Today, we are experiencing a global digital transformation that is accelerating with greater speed and voracity than ever before. As ongoing digitalization continues to expand and evolve all over the world, the deeply complex telecommunication systems enabling it will become even more robust and intricate in the coming years. As such, telecoms operators are in a position of perpetual demand for advanced methods, tools, and techniques that create data driven, intelligent, non-fragile systems for continual automation, evolution, and disruption. With widespread 5G and IoT integration on the horizon, which will further intensify the complexity of networks and operations as the number of functionality dimensions significantly increases, the task of managing and optimizing the 5G and IoT solutions of tomorrow far surpasses human capability and comprehension.
Currently in development by Ericsson, the world’s leading ICT provider, Machine Intelligence offers a means of reinventing network operations and redefining the operator product portfolio to create new business opportunities in 5G and IoT by accelerating digitalization on every front. Machine Intelligence enables operators like Ericsson to create a learning network for autonomous optimization while adequately handling increased network complexity, improving performance, and reducing costs.
By introducing Machine Learning and intelligent rule-based analytics directly into their products and software, operators can build network intelligence and in turn, increase efficiency by automating manual tasks while making products much more capable. Further, by introducing high levels of automation and using intelligence to predict, prevent, and handle operations without human intervention, operational expenditure is drastically reduced while customer service is markedly improved. These are easily achieved by MI implementation, which depends heavily on domain expertise, a thorough understanding of both Machine Learning and Artificial Intelligence, and virtually unlimited access to relevant data sets.
Machine Intelligence has the ability to transport intelligent functionality to enable RAN Edge Analytics and self-learning agents in the core network. In turn, it can help operators reduce in-field dispatches by up to 30 percent, reduce handover times between analytics cells as well as OPEX in RAN transport by 50 percent, and reduce energy consumption on the node level by 10 percent – all while making ICT solutions twice as easy to deploy. What’s more, Machine Intelligence can improve performance and lower operational costs by reducing AHT for service operations and engineering and improving resolution rates for operators, among other measurable improvements.
In order to implement Machine Intelligence, using the as a Service (aaS) delivery model will be key in data related business, as it decouples the data from the systems for greater speed, flexibility, and cost efficiency. To get started, operators can leverage leading vendors like Ericsson to identify opportunities where MI can be useful, beginning with already existing Artificial Intelligence and Machine Learning solutions and eventually moving into new solutions using operator data and analytics. From there, operators will need to run trials to measure performance and requirements for continued upkeep. Moreover, it will be vital for operators to break down silos in their own operations (such as RAN, Core, Edge, IT, etc.) and foster usage of data across the entire MI system. Finally, MI will also affect operators’ current OSS and BSS models, meaning they will need to work with the ecosystem to augment existing operations and create new processes.
With its Machine Intelligence strategy, Ericsson intends to move away from stand-alone on-premise analytics applications, which are typically slow and costly, in favor of adopting MI-based aaS offerings and implementations. Ericsson is taking a network first approach, leveraging its ICT expertise to build competence in Machine Intelligence as it moves forward with 5G and IoT development. As the complexity of network management and control continues to increase exponentially, MI will be instrumental for operators in building an optimal service management system, creating new effective personalized experiences for the customers and industries they serve, and ultimately growing their top line in the face of an unrelenting global digital transformation.
