3 minute read 9 Apr 2020
Close up of womens hands holding smartphone checking market data

How telco operators are optimizing their capex by applying AI

Authors

Tom Loozen

EY Global Telecommunications Leader

Fascinated by the positive impact of telecoms. Passionate musician. Enjoys educating himself on psychology, wine, sports, technology, arts and much more. Husband and father of three daughters.

Adrian Baschnonga

EY Global Telecommunications Lead Analyst

Lead Analyst with deep sector knowledge in technology, media and telecom, gained in professional services and business intelligence environments.

Contributors
3 minute read 9 Apr 2020
Related topics TMT Telecommunications AI

As companies face continued disruption, their current focus is on maintaining business continuity and employee well-being.

Now more than ever, our climate demands seamless delivery of services. 5G will ignite the next wave of the telecommunications industry, yet a smarter approach to capex planning will be vital, in this challenging new global environment, as operators aim to provide better customer experiences while also generating improved return on investment.

EY and B-Yond are developing solutions for operators as they invest in the next wave of digital infrastructure. By harnessing AI and automation in exciting ways, EY and B-Yond can help operators maximize their long-term value creation.

The growing mobile capex burden

A smarter approach to capex planning can deliver a better experience for customers, satisfy regulators and ultimately drive improvements in return on invested capital (ROIC). For this to happen, a new approach to capex planning is essential.

Realignment of people and workflows, with automation and artificial intelligence (AI) infused into network planning tools and processes, can drive a significant improvement in capex efficiency. A new mindset is needed for prioritizing the quality of network investment and its contribution to long-term value creation.

Global mobile capex development forecast

The capex challenge facing operators is increasing. Network
expenditure is forecast to rise significantly into the next
decade as operators deploy 5G and edge-cloud networks,
recasting their consumer and enterprise value propositions in
the process.
 

Capex planning in the 5G era

Capex planning in the 5G era needs to operate at unprecedented pace, scale and precision with seamless workflow across the entire organization on a continuous basis. Shared ownership and accountabilities are vital, as is a wider range of KPIs, including metrics relating to the quality of the customer experience and financial performance. Linear tools and scenario-based rules should give way to machine-learning capabilities supported by dynamic processes.

Planning considerations Traditional Fit to 5G
Ownership Fragmented Integrated
Operation Manual Automated
Accountability Network Business
Input variables 5 to 10 No practical limit
Tools  Linear Machine learning-based
Planning rules and process Static and scenario-based Dynamic
Planning horizon Over 12 to 18 months Continuous
Return on investment (ROI) Not modeled Prioritized
Workflow High-touch Seamless
Budgeting Network-driven Business-driven
Time to market Best effort Accelerated
KPIs Network Network, customer and financial

Key success factors for smart capex planning

Standardized AI technology can deliver a step change in agility and precision, by bringing together deep learning and dynamic modeling capabilities. There is no practical limit on the amount or type of data that can be modeled, paving the way for industrialized processes that are automated, transparent and self-governing. At the same time, vendor-agnostic platforms enable more flexible ecosystem relationships. With AI at the heart of capex planning, network planners can become better network investors.

Smart capex planning

Harnessing automation with AI-led planning tools can drive a better balance between people and processes that delivers shorter planning cycles, greater procedural transparency and more reliable decision-making. At the same time, operators can free up resources to upskill for higher-value functions.

Establish continuous planning

This is vital if network planners are to optimize their use of resources. Agile planning outputs — whether in the form of varied network scenarios or investment recommendations — demand ongoing oversight. In this way, operators will be able to make better decisions as they need them, as transformation road maps or competitor actions evolve in new ways.

Case study: US tier-1 mobile network operator

A US tier-1 mobile network operator wanted to improve its capital allocation for 5G rollout. It had been using rules-based models but was realizing the limits of using such tools for 5G networks. In addition, there was a mandate from business to improve customer experience.

B-Yond collaborated with the carrier, comparing the performance of traditional in-house capex planning tools with an AI-based planning tool for a fixed set of sites over a fixed planning period. Benchmark test results clearly demonstrated the advantage of AI-based smart capex planning over the current in-house tools.

Cell planning In-house tool Smart capex Improvement
Method Linear Dynamic Machine learning
Historical data 18 months 36 months Pattern-based
Validation period 4 months  4 months  Baseline
Correct triggers 478 550 72
Missed triggers 121 49 72
Excess triggers 213 76 137
Capex accuracy 59% 81% 37%
Underinvestment 25% 7% 72%
Over-investment 45% 11% 76%

The results were indisputable: a 37% accuracy improvement in capacity prioritization, 72% reduction in capacity degradation, and 76% reduction in unnecessary capex and opex (over-investment). But this is just the beginning; with AI, planning rules are dynamic and continuously evolving toward zero-error tolerance. As the AI-based tool continues to learn and evolve on the basis of network and traffic patterns, significant capex accuracy can be achieved.

Conclusion

Greater capex efficiency is no longer a desirable attribute, but a business-critical competitive differentiator in the 5G era. For this to happen, a major modernization of network planning activities is required on multiple fronts.

AI-driven tools can help provide improved customer experience and financial benefits while also streamlining workflows and upskilling resources. In this way, smart capex planning can act as a foundation for broader strategic goals.

Greater levels of automation will be the lifeblood of a new capex planning paradigm, one where business outcomes inform overhaul of tools, processes and roles. Legacy systems and processes will not disappear overnight, but AI-driven planning tools will provide a path to graceful attrition and retirement of aging systems without business or operational disruption.

Ultimately, network planning driven by AI and automation has the potential to deliver upside in three important ways by:

  • Optimizing capex on the basis of value to customers
  • Reducing customer churn
  • Streamlining business processes and increasing ROIC

Taken together, these benefits will determine which operators are able to thrive as opposed to surviving in the 5G era.

Summary

Greater levels of capex efficiency are vital for communications providers, particularly as they embark on the new wave of 5G investment. Network planning driven by AI and automation can deliver a better customer experience and drive significant ROI improvements especially given the current challenges faced across the   world today. With a smarter approach to capex in place, communications providers can thrive as opposed to simply survive in the 5G era. 

About this article

Authors

Tom Loozen

EY Global Telecommunications Leader

Fascinated by the positive impact of telecoms. Passionate musician. Enjoys educating himself on psychology, wine, sports, technology, arts and much more. Husband and father of three daughters.

Adrian Baschnonga

EY Global Telecommunications Lead Analyst

Lead Analyst with deep sector knowledge in technology, media and telecom, gained in professional services and business intelligence environments.

Contributors
Related topics TMT Telecommunications AI