A roadmap to zero-touch operations that entails automation, powered by AI and Machine Learning, to increase operational efficiency, handle virtual networks and meet the demand for high-quality customer service.
Today’s heavy reliance on connectivity drives customers to demand instantly delivered and high-quality innovative services, with zero tolerance for faults or downtime. In this context, CSPs face tough operational challenges, with heterogeneous vendor networks spanning multiple domains and needing constant, quick and minimally disruptive human intervention in Operation & Management.
Upcoming massification of IoT, new network paradigms like SDN/NFV and radically different 5G architectures are pushing the bar even further.
They demand a radical transformation of the operations that will be key to maintain costs down while still being able to cope with increasing demands for top service quality connectivity.
With this context in mind, CSPs face two major challenges:
- Automating configuration, detection, identification and resolution of problems in current legacy physical networks and environments is required to increase operational efficiency, contain OPEX and increase customer service quality as well as increase teams throughput by shifting people from repetitive low-value tasks to where they are most needed: strategical and preventive work.
- With 5G and SDN/NFV stepping up and bringing promises of delivering new and innovative use cases for new revenue streams, CSPs must prepare and implement roadmaps to deploy these new technologies. At the same time, they must keep pushing their architectures towards disruptive Autonomous Systems that are capable of extracting the full benefit of technology.
The Road to Zero-Touch Networks
To succeed in this emerging environment, CSPs need to embrace a long-term Digitalisation & Transformation strategy where their networks, operations’ processes and tools will be addressed within different vectors:
- Rethinking Operation & Maintenance processes to be more efficient.
- Applying innovative technologies to increase automation levels (Advanced Monitoring, Automation & Orchestration, AI and Machine Learning, RPA, SDN/ NFV and Cloud, among others).
- Shifting people’s and organisation’s mentality towards an agile automation culture with more focus and strive for proactivity and prevention instead of shear reactiveness to unexpected everyday situations. It’s crucial that everyone’s role in the FMO (Future Mode of Operations) is clear to avoid any friction or fear during the transformation.
For the automation journey to be successful, it’s critical to approach it as not just another technology challenge, but as a transformation journey instead.
The Journey from low automation to achieving full autonomous Network and Operation Systems encapsulates different levels and stages that CSPs must tackle:
Stage 0, Manual or Augmented Human (Manual) - Most Engineering & Operations processes are executed manually or with some automation from basic tools (e.g., scripts or excel macros).
Stage 1, Human Guided Automation (Process Automation) - Process automation is implemented and triggered by humans to execute time consuming, repetitive, error-prone tasks or others that need efficiency.
Stage 2, Human Assisted Automation (Cognitive Automation) - Automation empowered by AI and Machine Learning to take the process execution lead with AI-based decisions triggering automation processes. Here, humans assist automation only where it’s not technically possible for automation to execute specific tasks that complete the E2E process.
Stage 3, Autonomous System (Zero-Touch Network and Operations) - Fully autonomous networks and operating systems require minimal to no human intervention besides the definition of algorithms, policies and closed-loop orchestrations that define how systems and networks behave.
The Vodafone Group’s NOC mission is to monitor, control and ensure network availability across multiple geographies. To cope with a dynamic and unsteady environment, the company designed a long-term strategy for its OSS transformation.
Vodafone’s Cognitive Intelligence and Automation Suite for Network
The challenge was to improve cost reduction of 1st line Network Operation’s activities through cognitive intelligence and automation.
The solution autonomously surfaces relevant events detects meaningful incident patterns, diagnoses root-causes and runs automation flows, handling all the network incident management process from event to resolution.
In a nutshell:
- Live in 5 CSPs
- 80% goal for automated 1st line support activities
- 23 automation journeys
- 22% zero-touch
- 55M/month ingested events
- 98% monitored alarms
- 95% noise reduction
- 87% actionable alarms classified
Vodafone’s Next Generation Monitoring for IoT
The challenge was to detect, communicate and resolve service impacting anomalies faster to increase IoT’s service quality and customer satisfaction.
The solution centralised Service Usage and Resource Monitoring which leverages on Machine Learning to detect service incidents and correlate with most probable root-cause(s). Automation for increased efficiency, speeding-up anomaly detection, diagnosis, dispatching and healing.
Impact and Benefits:
- Increased detected anomalies rate
- Reduced false-positive anomalies
- Improved impact assessment
- Reduced MTTD: detected Anomalies in less than 5 minutes
- Communicated with customers before they complained
- Reduced MTTR: improved incident response