logo_header
  • Topics
  • Research & Analysis
  • Features & Opinion
  • Webinars & Podcasts
  • Videos
  • Event videos

Like it or not, data drives outside plant forward

TM Forum’s new report looks how communications service providers can apply artificial intelligence to data from outside plant to save money and improve customer experience. Report author Tim McElligott concedes this is a critical step forward but is wistful for days gone by.

Tim McElligott
08 Oct 2019
Like it or not, data drives outside plant forward

Like it or not, data drives outside plant forward

TM Forum’s new report Fusing outside plant data and AI to maximize capital investment looks how communications service providers (CSPs) can apply artificial intelligence (AI) to data from outside plant to save money and improve customer experience. Report author Tim McElligott concedes this is a critical step forward but is wistful for days gone by.

It would be unfair and incorrect to call outside plant a remnant, because it is still the primary element of communications networking – the foundation upon which all else stands. However, given the telecommunications industry’s focus on software, cloud and virtualization, outside plant too often feels like a remnant of what telecom once represented.

I, for one, love outside plant. I love seeing microwave towers in the distance rising above fields of blond, tasseled corn. I love being able to pick out a central office among all the glass and steel structures in the city as I pass through on a train.
Microwave radio tower in Norway, Illinois, circa 1987
I admire the independent streak of rural telcos who take extra pride in their field operations execution to show they are just as good or better than the big carriers. I love seeing techs up on poles and climbing into the underworld strapped with leather tool belts and a butt set banging against their hips. I also have a penchant for well-organized cable racks and feel comforted when disaster strikes and I see the flashing lights of a bucket truck ride up to face it.

So, forgive me for hesitating when asked to explore how AI can improve outside plant. Really? Does AI have to touch everything? Do we have to drag every, last blue-collar job into the geekdom of algorithms and big data?
As it turns out, we do. If service providers want to survive and stay competitive in the new digital economy, they must convince operations teams to accept help from data scientists and ethereal machines.

There is a perception that data scientists can’t possibly know what it is like to operate in the real world of paper-based engineering, manual labor, finicky networking equipment and indifferent weather, but they can. They may not know how to swap out a circuit board, but their algorithms can tell when it is about to go bad and based on this knowledge take corrective action before customers are affected. Data scientists may not know what it is like to stand atop a tower in the wind adjusting a radio, but AI systems can learn from performance patterns when the radio resources may be needed elsewhere immediately and change the configuration based on this knowledge.

Early days


CSPs are in the very early stages of understanding what AI can learn about usage and patterns and about the optimal design and configuration of a network or service and applying it. In a May 2019 survey of C-level telecom industry executives as part of its annual CXO Summit, TM Forum found that almost 20% have implemented AI or machine learning solution somewhere in the business and another 25% are in the processes of rolling out AI solutions.

The graphic below illustrates some examples from a 2019 Ericsson report of where CSPs are experimenting with, and to a lesser degree so far, deploying AI technology.
Three of the five areas have a direct connection to outside plant:

  • Optimizing performance, reliability and availability

  • Operational cost savings

  • Capital expense reduction


GIS helps automation


The operation and planning of outside plant rely heavily on geographic information systems (GIS). Randall Frantz, Founder of GIS consulting firm, RCF Consulting, explains that for most CSPs the work of planning, implementing and tracking network assets in outside plant is still a manual process that is labor intensive and not very accurate or consistent across organizations. However, this has started to change, thanks in part to advances in GIS technology and its extensibility to other applications for planning and engineering, operations support, and analytics.

Extending outside plant data to the rest of the business has also helped CSPs address the enterprise market with more accurate, timely quotes for services. AI and machine learning have helped operations teams provide quotes often without needing to make physical site visits. One example is 3-GIS, a US-based provider of geospatial asset management solutions, which applied machine learning to its planning software for a CSP in Dallas, Texas, running C-RAN technology with small cell and macro cell sites. The company needed 208 miles (335 Kilometers) of fiber routing for a project, which would normally take 15 weeks to design using traditional engineering applications. 3-GIS provided a routing plan within minutes.

To learn more about how AI and outside plant data can improve not just operational efficiency but the CSP business overall, read the full report. Then, the next time you see a lineman out on a rainy day, bring her a cup of coffee.