Nokia AVA AI-as-a-service
Injecting AI at scale on public cloud
Operating a telco network has always been complex. But with massive data growth and multiple layers of network technology, tomorrow’s IoT and 5G networks will be more complex than ever. With pressure on headcount, many operators are already reaching the limits of what their human engineers can achieve. That’s why AI and analytics, with help from automation, is vitally important for the future of network operations.
Nokia AI-as-a-service is designed to help telcos overcome the challenges to improve operational effectiveness, productivity and enhance the customer experience. Most operators are seeing flat or falling revenues. AI and analytics can potentially help reduce churn and incubate new sources of revenue.
The Nokia AVA Telco AI Ecosystem provides Artificial Intelligence solutions that can be delivered through the cloud with Microsoft Azure or other public options. The solutions enable Communications Service Providers to automate network operations and service assurance, cutting costs, increasing agility and boosting subscriber experience. Our solutions are delivered as a service, so there’s zero upfront CAPEX.
Key challenges for effectively implementing AI in telecoms
The top 6 key challenges that CSP face, based on detailed interviews with 9 leading CSP Chief Data Officers/ Strategists and survey of 50 CSPs worldwide, conducted by STL Partners an independent research company.
Data quality is a key issue, often consuming
> 60% of efforts and time in AI projects.
Data science skills
55% lack the right data science and machine learning productization skills as pure data scientists are difficult to find, recruit and retain.
80% of AI POCs fail to reach live deployment - this can be poor strategy, lack of central co-ordination or other people, process, technology issues.
77% of CSPs struggle with data storage
and data fit for exploitation
23% of telco's are concerned about Security when it comes to AI.
How does Nokia AVA work?
Nokia AVA AI as a service allows CSPs to deploy AI use cases quickly and securely, completing data set-up in four weeks. By integrating Nokia’s security framework with Azure’s digital architecture, deployment is much faster while concerns around data sovereignty and security are eliminated and global regulations respected. Azure’s platform enables Nokia’s AI deployments to be delivered fast and enables CSPs to scale fast across multiple clusters.
After the initial data set-up, CSPs can deploy additional AVA AI use cases within one week and ramp-up, ramp-down resources as needed within one day across multiple network clusters. The Nokia security framework on Azure ensures data is segregated and isolated to provide the same level of security as a private cloud.
How do we drive successful AI deployment
1st telco AI deployments on public cloud worldwide
Security framework for ‘any’ cloud deployments
Mature Telco AI library
Outcome based model – no gain, no fee
Scales fast according to number of use cases and size of network
Go live from 9 months to 4 weeks as no need to deploy hardware
Nokia AVA AI-as-a-service use case library
Reach sustainability targets
- Field services optimization
Up to -5% in emergency field ops OPEX
- Energy efficiency
Up to -20% OPEX for energy
Increase NPS scores and revenue from new services
- Predictive video analytics
-59% buffering and -15% less play delays
- Service management
Double digit NPS boost & triple digit productivity gain
- Spectral performance
Up to + 17% efficiency
- Geo data as-a-service
Better prioritization / avoiding CAPEX
Reduce OPEX and increase productivity in operations
- Cell site degradation prediction
Up to 7 days in advance with 85% accuracy
- KPI degradation detection
100% success in detecting sleeping cells near real-time
- Automated assisted triage
OPEX⬇ eliminating manual processes
- RF coverage optimization
Up to -80% cost related to manual drive testing
- Intelligent alarm clearing
Up to -82% in mean time to recover