Blackouts are so routine in Nigeria that neighbourhoods measure time by the drone of generator sounds.
The Country’s national grid had experienced 12 collapses so far in 2024, a troubling trend that reveals the fragility of the country’s energy infrastructure.
A World Bank assessment puts the yearly cost of unreliable power supply to nearly $29 billion, roughly 5% of the country’s national output.
But behind the outages lies another quiet drain: According to NERC’s 2024 quarterly report, distribution companies let 35.22% of the electricity they receive leak away through worn lines, billing gaps, and outright theft, wiping out ₦139 billion in a single quarter for the DisCos.
But Nnadozie Odinaka, with over half a decade of experience analyzing energy topics across continents has an idea to plug these leakages.
His proposal relies on artificial intelligence that sifts through the flood of data from smart meters, feeder sensors, and transformer monitors to pinpoint exactly where energy is disappearing.
A tampered meter, an overheating coil, or a sagging line each leaves a unique fingerprint in the data.
Software can spot that pattern in seconds, send an alert to the control room, and dispatch crews before the first kilowatt is lost.
Evidence from abroad shows the idea works. Nnadozie references Delhi, where utility company, BSES used AI on millions of meter readings to push network losses down to about 7% and flag power theft within minutes.
In Chile, Enel’s predictive algorithms recovered billions of pesos by detecting illegal taps and prioritising repairs where they mattered most.
Lagos engineers are already piloting transformer health sensors that stream real time readings into an analytics hub, cutting fault detection time and proving that local data streams can feed intelligent tools.
Nnadozie sketches three practical moves. First, accelerate smart meter rollout and mandate that every new device transmits encrypted readings to a shared analytics platform.
Second, create an innovation sandbox at the regulator so distribution companies can test theft detection algorithms without years of paperwork.
Third, issue a loss reduction bond that channels the savings from each percentage point of recovered energy into financing the next wave of sensors and software.
He believes the payoff would be immediate. Recovering even half of today’s technical and commercial losses would unlock more than 2 000 megawatts of effective capacity, equal to two large gas plants, without pouring a single cubic metre of concrete.
That extra headroom would let factories switch off diesel sets, hospitals run life saving equipment, and students read under steady light rather than candle glow.
Nnadozie Odinaka insists that the data we already collect can be turned into reliable power with code and commitment.
“Every smart meter is a reporter on the grid,” he says. Listen to them, and we can keep the lights on while bigger projects catch up.”
His call invites power managers, regulators, and technology firms to join forces now, proving that a digital fix can deliver real electricity long before the next plant comes on line.
and then