Login
Register
Read more about how Taiwan Mobile implemented an open telco OSS framework, based on TM Forum's Business Process Framework and an AIaaS platform.
Taiwan Mobile transforms its OSS with an AI ecosystem
Who: Taiwan Mobile & Nokia
What: Transformed its OSS with automation and AI analytics to deliver better insights and manage complexity as 5G/IoT looms
How: Implemented an open telco OSS framework, based on TM Forum's Business Process Framework and an AIaaS platform
Results: A key thing to understand about digital transformation is that it’s a journey – and sometimes journeys can take several years to complete, depending on the complexity and age of the organization being transformed.
Taiwan Mobile may not be old by incumbent standards, but it was one of the first newly licensed operators to emerge from Taiwan’s telecoms liberalization and launch services in 1997. Taiwan Mobile has grown its business considerably since then and ranked as the market’s second-largest mobile operator after incumbent Chunghwa Telecom. However, by the 2010s, the operator found itself saddled with a complex multi-vendor matrix of services and network domains across its 2G, 3G and 4G networks. Taiwan Mobile was struggling from an operations standpoint to handle the increasingly massive amounts of data its networks were generating. In 2015, Taiwan Mobile contracted Nokia to transform its operational support systems (OSS) in a multi-stage journey that is still ongoing. So far, the ‘Smart Network Operations Transformation’ journey comprises two overall phases, with the first focused on the customer experience, and the second driven by big data and artificial intelligence (AI). What’s especially noteworthy about the latter stage is that it incorporates AI into Taiwan Mobile’s OSS via Nokia’s platform-based AI ecosystem – AI-as-a-service (AIaaS), if you will.
Before getting into the details of that, it’s worth reviewing the first stage of the OSS transformation journey, which ran from 2015 to 2018. Before the project started, Taiwan Mobile’s OSS applications were standalone and siloed, resulting in data being collected, processed and stored in multiple data storage silos with lots of overlap. This limited the operator’s ability to perform critical operations functions such as root cause analysis and customer impact analysis. So, for Phase 1, Nokia provided Taiwan Mobile with end-to-end network and service assurance solutions in order to gain better insights into network problems, their root causes and impact on the customer experience. Nokia helped consolidate Taiwan Mobile’s 2G/3G/4G/fixed network OSS and probed data sources to build a centralized OSS data lake. This transformed Taiwan Mobile’s standalone, siloed applications into an open telco OSS framework with consolidated umbrella applications sharing data through open APIs. Tom Koh, Senior Vice President and Chief Technology Officer, Taiwan Mobile, says the open telco OSS framework adopts TM Forum's Business Process Framework (also called eTOM), part of the Open Digital Framework.
Nokia started with the resource management and operations layer to address network resource quality analysis and problem resolution. The solution sticks consolidated and re-usable data assets in the OSS data lake to support umbrella apps for fault management, configuration management and performance management. Then the service management and operations layer was added to address voice over LTE (VoLTE) service quality monitoring and management. Finally, in the customer relationship management layer, they implemented impact analysis for individual and group level customer experience. Koh adds that the OSS solution also incorporates the IT Infrastructure Library (ITIL) framework for IT service management, focusing on operations assurance. “The basic ITIL methodology of service design, service transition and service operation is very useful to us mainly for request management and change management. We can also link it with our Agile development,” he says.
With this foundation firmly established, the second stage of the OSS transformation journey commenced in 2018, at which point Taiwan Mobile was weathering disruptive changes in its mobile business. Between the maturation of LTE and VoLTE, and the approaching advances in 5G and IoT, the operator’s network operations complexity and costs were increasing at an alarming rate. Even with the open OSS framework in place, Taiwan Mobile needed a more advanced intelligence and automation framework to optimize daily network operations, enhance end-to-end service quality and reduce its total cost of ownership for network and customer experience management. That’s where AI comes into the picture– or, more accurately, an AI ecosystem. Taiwan Mobile was already exploring AI and machine learning (ML) as an automation solution, but the problem with AI/ML in 2018 (as well as now) was that it was still a nascent market. Consequently, AI/ML specific experience and expertise was not readily available. Even if it was, the AI/ML solutions market was a highly fragmented mishmash of algorithms, models, tools and stacks, which made it difficult to choose the right tools and solutions that could be rapidly integrated with the operator’s existing applications and systems. It also meant that AI/ML use cases were being developed for individual application silos, which was difficult to manage and optimize at the enterprise level.
By contrast, Nokia’s AVA Telco AI ecosystem offered something different – an AIaaS platform that provides telco-specific full life-cycle management for AI/ML-driven applications in the network operations process. Based on a container and microservice architecture, the AVA platform is open, decoupled and ready to integrate with an operator’s big data ecosystem and applications. AVA also comes with an out-of-box Telco AI/ML use-case library with patented algorithms by Nokia Bell Labs and Global Services, which Koh of Taiwan Mobile says is a major time-saver: “Productizing an AI/ML use case requires long development cycles and an intensive testing process in the production network. We need out-of-box AI/ML use cases that can be immediately integrated in network operations and optimization processes to rapidly generate value.” The platform also provides multiple development templates and codeless graphical tools for different type of users, with a in-built DevOps process to enables continuous integration and continuous delivery (CI/CD) for agile and rapid end-to-end deployment. This turned out to be a key feature to enabling agile and efficient execution of the project, says Koh. “In the OSS domain, there are too many vendors, equipment, technologies and interfaces, while customers’ requirements are also different and often changing,” he explains. “We need to fully understand the customer’s pain points, quickly develop and test new functions and deploy in the actual operation system.” Taiwan Mobile also took advantage of Nokia’s AI consultancy and training services, with over 100 employees acquiring new AI/ML skillsets to apply in various areas of their work.
Taiwan Mobile’s ‘Smart Network Operations Transformation’ journey is now in its fifth year, and between the two stages so far, its 4G and 5G network operations and service management has improved significantly since 2015:
All up, the transformation project has enabled Taiwan Mobile to set up a smart OSS and AI ecosystem with full network operations monitoring, thus highly improving operational efficiency. Meanwhile, Taiwan Mobile says its ‘Smart Network Operations Transformation’ journey is still ongoing, and expects it to yield even more benefits in the future.
