Walk into the operations centre of almost any tower company today and you’ll see what looks like a fully digitised business. Dashboards everywhere. Green ticks, red alarms, dense scrolling streams of telemetry from thousands of sites. On the screen, the network looks firmly under control.
The engineers know better.
Ask one of them what they actually do when a ‘generator failed to start’ alarm fires at three in the morning, and the answer is rarely “look at the dashboard and act.” It’s usually some version of “send the field technician out, ask what they’re seeing, and check the last refuel record by hand.” The dashboard is there. The data is there, but the trust often isn’t.
These are not operators failing to use the tools available to them. They are experienced professionals who have learned - correctly - that their telemetry cannot always be relied upon. That hard-won scepticism is costing the industry enormously, and the answer is not better instincts. It’s better data.
PowerX ran an AI-readiness assessment across hundreds of live tower sites in Africa, processing tens of million telemetry records. We weren’t just auditing power assets and their performance - we were auditing the reliability and robustness of the data those assets were producing and the impact on the operational efficiency of the hybrid power systems in place. The results speak for themselves.
These are just a flavour of the data or operational issues observed. The broader point is this: network operators are making capital and operating decisions - refuelling schedules, battery replacement plans, solar payback cases, hybrid control logic - on the basis of measurements that are frequently wrong, contradictory, or absent. The engineers working with this data know it. They’ve adapted their workflows around it. What they haven’t had is a way to fix it systematically, at scale.
The temptation right now is to bolt AI on top. Almost eight in ten companies globally have deployed generative AI somewhere in their business - and more than eight in ten of those report no material impact on earnings. The reason isn’t that the models are bad. It’s that they were pointed at workflows whose underlying data was already broken.
Tower networks risk walking into the same trap. Feed unreliable telemetry into an optimisation engine and it doesn’t fail quietly - it accelerates errors and recommends interventions with misplaced confidence. The competitive advantage in the next five years won’t go to the operator that bought the smartest model. It will go to the one whose models are running on a data foundation you can trust.
Fixing the data layer isn’t a glamorous strategy. But the financial case is immediate, and it doesn’t require anyone to believe in AI to be persuaded.
The PowerX platform starts by doing what most monitoring systems don’t: standardising telemetry across all RMS hardware in an agnostic way, applying deep discipline to data handling at the edge with strong hardware and firmware partnerships, validating and reconciling the telemetry itself. Continuous quality scoring at signal, asset, and site level. Cross-checking measurements against expected physical behaviour. Separating genuine operational issues from sensor faults and configuration drift. Where gaps exist, synthetic data techniques maintain analytical continuity. Where the data is reliable, machine learning surfaces patterns that are invisible to manual review - early battery degradation, hybrid system misconfiguration, solar underperformance - and prioritises them by financial impact.
When operators can trust their data, the results follow quickly. Generators stop running when they don’t need to. Batteries stop discharging while grid power is available. Fuel deliveries reconcile with consumption, and discrepancies become actionable. Batteries due for replacement get flagged before they affect SLA performance. Across a five-thousand-site network, even modest gains here run into the millions.
And then - once the foundation is solid - AI continues to learn and deliver continuous improvements. Each clean signal sharpens the models. The operators who start this work earliest will be the ones whose systems are years ahead by the time everyone else is catching up.
Tower operations teams are skilled, experienced, and adaptable. What they deserve are tools that give them data they can actually trust. That’s where transformation begins.
Read the full PowerX whitepaper, Data Crisis in the Tower Industry.