September 27, 2023

What does the Perfect Tower Look Like?

Justin Head

Founder, Executive Vice-Chairman

For this week’s blog, I thought we could try a little thought experiment.

Let’s imagine for a moment that your telco has completely unlimited resources – an endless supply of free everything – how would you maximise the efficiency of your network?  

Well, since a network is – by definition – an aggregation of individual towers, let’s start there: with a single tower. What could we do to make this one tower run at its absolute best, providing the best end user experience, while reducing OPEX costs and CO2 emissions.

At present, this tower has its own operational parameters and is configured to switch between different power sources , shifting between grid, diesel, battery, and possibly solar to meet organizational goals – whether they’re reducing CO2 emissions, minimizing costs, maintaining service levels, cutting service calls etc. But as we’re all too aware, tower operation is fluid, dynamic and dependably unpredictable.  Conditions change, loads increase and decrease, equipment degrades and fails, anomalies in one component can have unforeseen impacts on the entire tower and a host of other, unforeseen issues can throw a wrench into the best thought-out preprograming at a tower location. Returning to our thought experiment then, how would unlimited resources solve the problem of keeping this tower operating at peak efficiency, as well as meeting the goals of the organization as a whole?

It may be a business cliché (and the thing about clichés is that they have some truth at their core), but an organization’s biggest asset really is its employees – or more specifically, the intellectual capital that these employees embody. It’s the analytical, problem-solving, creative, and decision-making power of employees that drives business success.  So, if we want our imaginary tower to operate at peak efficiency, we should have a team of these trained, experienced professionals on-site 24 hours a day, tweaking and adjusting its operation in real-time.

I was brought to this thought experiment by one of our customers in South Africa, Andre Herbst, the CEO of MEGMAR. MEGMAR has been at the forefront of the push for artificial intelligence in tower networks, and Andre has been a vocal advocate of the benefits of AI automation.  He described the PowerX platform like this:  

It’s like having an accountant standing at the tower site, along with an operations manager, a technical guy, a sustainability consultant, a cell provider customer service rep and a data analyst all making decisions together to have this one little site in the middle of nowhere operating at the cheapest, most efficient, most environmentally friendly level possible.

Across a network, data science, AI/ML means you can have this team of virtual experts at every single tower, 24 hours a day, 365 days a year – scaling up the benefits. Take for example a typical function such as the switch from using battery to generator power. The controller is preprogrammed so that when the batteries reach a certain level (including a buffer), the generator kicks in. Because it’s preprogrammed, this happens every day following the same trigger event. But with a  virtual team of experts on-site, the data science power an AI ‘operations manager’ looks at the system as whole and decides based on the weather forecast, expected use and  other relevant factors, there’s enough power to delay that generator for 45 minutes. That doesn’t sound like much, but 45 minutes per day, over 12,000 sites, over the course of a year, has huge financial and carbon emissions implications.  

And this is just one event.  Each tower experiences up to hundreds of these events every day. With data science, AI/ML automation every one of these events undergoes the scrutiny and analysis of this virtual team working together 24 hours a day to iron out even the smallest of anomalies or inefficiencies.

Just like their human counterparts, the data science team of virtual on-site experts learns the idiosyncrasies of individual towers based on their components, location, and unique operational challenges. Scale these gains across an entire network, and the cost-savings, emission reductions and predictive maintenance benefits can be profound.

In world of unlimited resources, you could staff every tower with a team of experts making sure every aspect of the tower’s operation was maximised to the full. With artificial intelligence tools, that’s exactly what you get – without the need for unlimited resources.

related Blog POsts