Automation must be intelligent to make services more agile and dynamic
Updated: Apr 23, 2020
Automation has been the central tenet of the OSS/BSS vendor business from the beginning, taking existing manual processes in a telco’s operations and business and code software that will run those processes automatically without being watched over by an expensive human. The increasing complexity of digital networks and service types means that process automation like that found in legacy systems is becoming increasingly flawed as more and more decision-making is required in workflows and orchestration schemas, contextual decision making with reasoning and intelligence similar to that of a human operative. The difference being that this must be human-esque intelligence taking into account thousands of rapidly changing variables and processing many thousands of these decisions per second in some cases. CSPs have turned to artificial intelligence (AI) and machine learning (ML) technologies to address this challenge, but it is not just a matter of layering an AI engine over the OSS/BSS data, the complexities require the intelligence to be both distributed and native in the systems themselves and centralized to make overarching optimal conclusions.
Hyperscale transactions in new service models
SD-WAN is great example of making significant operational gains by deploying an intelligent automation strategy from the ground up. An SD-WAN solution operating on a microservices-based, cloud-native platform, with zero-touch automation across IT and network certainly qualifies as a cutting edge CSP service, but the integration of analytics empowered with AI and ML for network insights and intelligence is the game changer. The challenge then lies in translating these advantages into a platform that can operate at hyperscale levels, but operationally, the effects can be significant:
· Huge cut in provisioning process times
· Reduction in the manual input data error rate
· Increased efficiency in service fulfilment processes
· Improvements in fault resolution
How does this ‘intelligence’ manifest itself in the digital operations platform?
Data collection and analysis: is the primary discipline for any analytics system in the telecoms industry, sourcing vast amounts of data from customers, the network, IT systems and anywhere that usable numbers can be found. Analytical analysis of this data is now a mature and well-understood data science, but shaping that analysis to be telecoms service specific is a complex business. We are seeing strong partnerships in the industry between CSPs and their vendor partners to hone this art and produce the most actionable insights possible.
Building transformative steps from those conclusions: is one of the most difficult steps in the, but this is where AI and ML can shine by producing software-defined processes to not only radically improved CSP operations but improve the BDA’s own processes. This is where additional agility and dynamicism can be introduced as the analytics intelligence grows and is able to take on increasingly complex decision making exercises both within individual software modules and as an overall end-to-end process chain.
Informing the design and deployment of future digital services: is the long-goal for intelligent big data analytics algorithms. This concept sits neatly with the idea of a modular software architecture that is constantly being refined and improved. In this next-generation model, the software architecture and standards themselves are partially designed by the software.
The resultant operational capabilities will have a significant impact on customer experience
The long-term goal for CSPs will be a hugely improved operational framework, with reliable intelligence, which is very specific to their current business requirements, but agile and dynamic enough to pivot onto new opportunities without the traditional upheaval. Processes such as order-to-cash and cash-to-care are liberated from being static flows to dynamic, governed logic trees, enforcing optimal outcome. The net result of this fundamental shift in thinking is that excellence in customer experience is forced through by the software itself along with a cultural change within the operator.
The cultural change of CSP gets a lot lip service at industry events, but it is becoming clear that the ability to change thinking in the business is not equal to the appetite. Therefore the sensible way to make this change going forward is by making the desired way of doing things, the only way of doing things. AI and ML empowered analytics provide that leadership by eliminating classic telecoms problems at scale.