With a cloud-only monitoring platform, this site is blind. The anomaly will not be detected until morning — if at all. The operator in Lagos opens their dashboard, sees a data gap, files it under "connectivity issues." By the time a technician reaches the site, the damage is done.

Now imagine a different architecture. A Raspberry Pi sitting next to the inverter runs a quantized anomaly detection model locally. It reads BMS data over Modbus, detects the thermal deviation within minutes, triggers a local alert, and adjusts the charge controller to reduce stress on the affected cells. No internet required. When connectivity returns, the event log syncs to the cloud.

This is not a thought experiment. This is technically and economically feasible today for mini-grids across Africa.

Cloud-Only Monitoring Is a Structural Flaw, Not an Edge Case

Let me be direct: if your monitoring platform requires a persistent internet connection to function, it does not work in rural Africa. It works in urban Africa. It works in the demo. It does not work where most mini-grids actually operate.

The NCC reported in 2025 that only 23% of rural communities in Nigeria have internet access, compared to 57% in urban areas. The national broadband penetration figure crossed 50% in late 2025, but it masks a severe rural-urban divide. The communities where mini-grids are most needed — the ones without grid connections — are overwhelmingly in that unconnected 77%.

And "access" is generous. Access means a 2G or 3G signal that drops regularly, has latency measured in seconds, and costs operators real money per megabyte. This is not a temporary gap that 4G rollouts will fix next year. Operators reported over 19,000 fibre cuts and 3,000 cases of equipment theft in the first eight months of 2025 alone. The connectivity problem is structural, and any monitoring architecture that treats it as an exception rather than the default is architecturally flawed.

The consequences cascade. Data gaps mean anomalies go undetected. Undetected anomalies become equipment failures. Failures at unmanned rural sites become extended outages, revenue losses, and community trust erosion. And when compliance reporting requires continuous operational data — as DARES increasingly expects — those gaps become regulatory problems too.

What Changed: The Cost Curve That Makes Edge AI Viable

Two years ago, deploying AI at the edge of a mini-grid was technically impractical. You could send all data to the cloud (requires internet that does not exist), install dedicated edge hardware ($500-2,000 per unit, uneconomic for a site generating $3,000-5,000/month), or skip AI entirely. Most operators chose the third option.

What changed is the convergence of three technology trends that individually were incremental but together are transformative.

First, inference frameworks matured for constrained hardware. TensorFlow Lite and ONNX Runtime now enable meaningful ML inference on devices with limited compute. TFLite Micro runs on microcontrollers with less than 1MB of memory. Benchmarks show TensorFlow Lite delivers 3-4x inference speedups over standard TensorFlow on ARM processors, putting a Raspberry Pi in competition with dedicated edge hardware costing ten times more.

Second, model quantization became reliable. Converting model weights from 32-bit floating-point to 8-bit integers (INT8 quantization) shrinks models by 4x with less than 1% accuracy loss on most tasks. INT8 delivers up to 16x performance-per-watt improvement compared to FP32. A battery degradation model that required a GPU two years ago now runs on a $75 single-board computer.

Third, the hardware got cheap enough. A Raspberry Pi 5 with 4GB RAM costs $75 — quad-core ARM Cortex-A76 at 2.4GHz, Wi-Fi, Gigabit Ethernet, enough compute for multiple ML models simultaneously. For context, the NeurIPS 2024 Edge-Device LLM Competition pushed researchers to run large language models — far more demanding than our time-series models — on comparable hardware.

The economics now work. A $75 edge node amortized over three years costs roughly $2 per month. Research published in January 2025 found that hybrid edge-cloud architectures yield energy savings of up to 75% and cost reductions exceeding 80% compared to pure cloud — savings that compound across a fleet of dozens or hundreds of sites.

What Actually Runs on a $75 Device

Let me be specific, because vague claims about "AI at the edge" help no one.

A Raspberry Pi running quantized models can handle:

• Battery health prediction — LSTM models analyzing charge/discharge curves, cell voltage deltas, and temperature patterns to predict degradation. This catches the 2am thermal anomaly before it becomes a failure.

• Inverter anomaly detection — XGBoost classifiers trained on inverter telemetry (efficiency curves, THD, temperature, error codes) that flag deviations from normal operating patterns. The edge node knows before the cloud does.

• Demand forecasting — Prophet-based models using historical consumption patterns and day-of-week seasonality to predict load, enabling smarter battery dispatch and reducing diesel runtime at generator-backed sites.

• PV yield estimation — Lightweight models correlating irradiance, panel temperature, and historical output to detect underperformance. A sudden drop relative to conditions might mean harmattan dust accumulation or a wiring fault.

These models are small — a quantized LSTM for battery health runs in single-digit megabytes. Inference takes milliseconds. And critically, they operate autonomously. When the internet drops, the site continues monitoring, detecting, and responding. When connectivity returns, the edge node syncs compressed event logs to the cloud for fleet-level analytics and model retraining. Each layer does what it does best.

The Integration Layer: Why Victron's Open Ecosystem Matters

Hardware intelligence is only useful if it can talk to the equipment.

Victron Energy's ecosystem stands out in the Nigerian market for a reason beyond hardware quality. Venus OS — the Linux-based OS running on Victron's GX devices — is open source. It exposes system data over MQTT and Modbus TCP, with full read and write capabilities. The community has built custom Python drivers that run directly on GX devices, integrating third-party sensors and controllers.

An edge compute node reads real-time data from the inverter, charge controller, and BMS over the local network — no internet required, no proprietary API gateway, no per-device licensing. It writes back too: adjusting charge parameters based on ML model outputs.

Compare this to closed platforms where data extraction requires a cloud API call — which requires internet that may not exist. For fleets running SparkMeter or FusionSolar, the integration patterns are similar: local Modbus or API access, data aggregated at the edge, cloud sync when available.

The Technician's Experience: Offline-First Mobile Apps

Edge intelligence is not just about the site controller. It extends to the human layer — the technicians who maintain these systems.

A technician driving to a site in rural Nasarawa should not arrive to discover their monitoring app cannot load because there is no signal. Offline-first Progressive Web Apps (PWAs) treat the network as an enhancement, not a prerequisite. Work orders, diagnostic data, equipment history, and alert details cache locally. The technician reviews site status, logs maintenance, and captures readings entirely offline. When they drive back through coverage, the app syncs automatically.

This architecture is well-established in healthcare (community health workers) and agriculture (extension officers with field apps). But it remains surprisingly uncommon in mini-grid O&M platforms, where the assumption of connectivity persists despite daily evidence to the contrary.

A Design Philosophy, Not a Feature

Offline-first is not a feature you bolt onto a monitoring platform. It is a design philosophy that shapes every architectural decision.

If your telemetry lives only in the cloud, you have a single point of failure that correlates with the exact conditions where monitoring matters most. If your alert engine runs only in the cloud, every connectivity outage is also a monitoring outage. If your technician app requires a live connection, it fails precisely when the technician is at the site where they need it.

The principle is straightforward: if your system does not work offline, it does not work in rural Africa.

A platform combining edge ML inference, autonomous local operation, cloud fleet analytics, and offline-first mobile tools would have been a PhD research proposal in 2022. In 2026, it is a buildable engineering project.

What remains is for the sector to demand it. Operators evaluating monitoring platforms should ask one question: what happens at my site when the internet goes down?

If the answer is "we wait for connectivity to return," that platform was not designed for this market.

What would you add to the list of capabilities a mini-grid monitoring platform must support offline? I would like to hear from operators, O&M teams, and hardware integrators — what does your site need to do when the internet drops?

Sources referenced in this article:

• NCC: Only 23% of rural communities have internet access in Nigeria (Vanguard, 2025)
• Nigeria's broadband penetration crossed 51% in 2025 (Independent)
• Quantifying Energy and Cost Benefits of Hybrid Edge Cloud (arXiv, January 2025)
• NeurIPS 2024 Edge-Device Large Language Model Competition
• Benchmarking Deep Learning Models on Edge Computing Devices (arXiv, 2024)
• LiteRT-Optimized INT8 LLM for Raspberry Pi 4 Deployment (ICCV 2025)
• Performance Monitoring of African Micro-Grids (NREL, 2020)
• Benchmarking Africa's Minigrids Report (AMDA, 2024)
• Edge Computing Solutions for African Telecommunications (Telecom Review Africa)
• AI Model Compression: Pruning and Quantization Strategies (Promwad, 2025)

DARES — the $750 million Distributed Access through Renewable Energy Scale-up project — is fundamentally a performance-based programme. The World Bank structured it so that mini-grid developers earn grants based on verified results. You commission a site, connect customers, prove reliable service for 12 months, and the money flows. Up to $600 per verified connection, with a significant holdback contingent on proven reliable operations. The Independent Verification Agent visits your site. The data either backs your claim or it does not.

This is pay-for-outcomes logic. Incentives are aligned, capital flows toward results, and underperformers are filtered out naturally.

And yet every piece of operational software serving the mini-grid sector in Nigeria charges flat monthly subscriptions. Per-site fees. Per-seat licenses. Annual contracts denominated in dollars. The programme itself is performance-based. The tools that operators depend on to deliver that performance are not. They get paid whether the operator succeeds or fails.

Something is broken here.

Why SaaS Pricing Fails Nigerian Mini-Grid Operators

Let me walk through the economics that most software vendors either do not understand or choose to ignore.

A mini-grid developer deploying under DARES has typically borrowed capital at Nigerian commercial lending rates — which hit 29.79% on average in January 2025, with some manufacturers paying 35%. They have invested that capital into site construction before receiving performance-based grants. The grant structure is milestone-driven, with the critical final tranche held back pending 12 months of demonstrated reliable service.

The developer is cash-negative from day one, servicing expensive debt while waiting for grant disbursements that depend on verified performance data.

Now layer in the naira. Nigeria's currency depreciated 41% against the dollar in 2024, hitting an average of ₦1,479 per USD. Annual inflation surged to 31%. Operational costs for businesses rose 67% year-on-year. Revenue is collected in naira from rural communities. Grant disbursements flow through processes that can take months.

Into this environment walks a monitoring platform saying: "That will be $150 per site per month, payable annually in advance, in dollars."

For an operator running 30 sites, that is $54,000 a year in fixed software costs — before the platform has demonstrated it will save a single naira. At current exchange rates, software costs can eat 3-5% of gross revenue. For a sector where margins are already razor-thin, that is an existential line item.

The deeper problem: the operator cannot evaluate whether the software delivers value until they have used it for months. But the pricing model demands commitment before proof. The vendor's revenue is decoupled from the operator's outcome.

This is exactly the misalignment that DARES was designed to eliminate from the grant structure. Why are we tolerating it in software?

What Diesel Waste Actually Costs

Let me make the savings concrete, because this argument lives or dies on the numbers.

The SEforALL Mini-grid CAPEX and OPEX Benchmark Study (August 2024) — covering Nigeria, Burkina Faso, and Sierra Leone — found that fuel constitutes 51% of ongoing operational expenditure for diesel-hybrid mini-grids. For the typical Nigerian hybrid site spending ₦500,000 per month on diesel, that is the single largest controllable cost centre.

AI-optimised diesel dispatch — intelligently scheduling genset runtime based on load forecasting, battery state of charge, and solar irradiance prediction — captures 15-30% more value than rule-based systems. Stem demonstrated this at utility scale with their Athena platform across 30 GW of assets. The technology is not theoretical.

For that ₦500,000/month diesel site, 15-30% optimisation means ₦75,000 to ₦150,000 in monthly savings. Per site. Every month.

Scale that across a fleet. DARES targets 1,350 mini-grids. Even if half run diesel-hybrid configurations, the aggregate diesel waste from suboptimal dispatch runs into billions of naira annually. Money burned — literally — because dispatch decisions are made by timer-based rules instead of intelligent systems.

O&M more broadly represents 33% of total tariff cost build-up. Equipment failures that go undetected for days compound into premature battery degradation, shortened inverter lifespans, and cascading costs from reactive rather than predictive maintenance.

The value is there. The question is who captures it, and how.

What If Operational Software Was Priced Like DARES Grants?

Imagine a different model.

An operational platform connects to your fleet — Victron VRM, SparkMeter Koios, FusionSolar. It monitors your sites, optimises diesel dispatch, automates your Odyssey reporting for DARES compliance, detects faults before they cascade. And you pay based on verified outcomes. A percentage of measured diesel savings. A fee tied to accelerated grant disbursements. If the platform does not save you money, you do not pay.

This is not radical. The energy sector has precedents.

Omnidian guarantees that every solar asset under its management will produce at least 95% of projected output — and pays the difference if it falls short. Revenue tripled between 2022 and 2024. The model works because Omnidian only profits when asset owners profit.

ESCOs have operated on shared-savings contracts for decades. The ESCO finances the efficiency upgrade and splits measured savings with the customer. If savings do not materialise, the ESCO absorbs the loss. The IEA documents these models as proven globally.

Schneider Electric's EaaS and Bloom Energy's zero-upfront power solutions convert capital expenditure into performance-guaranteed fees. Customers pay for outcomes, not for licenses.

The thread connecting all of these: the vendor's revenue depends on the customer's outcome. This is precisely how DARES structures its grants. Why should the software layer operate on a different philosophy?

Addressing the Objections

I know what the sceptics are thinking.

"How do you verify savings?" The same way DARES verifies connections — with data. Baseline diesel consumption before optimisation versus measured consumption after. Grant disbursement timelines before automated compliance reporting versus after. Mini-grid systems generate telemetry continuously. The measurement infrastructure already exists. A vendor whose revenue depends on verified savings has a stronger incentive to ensure metering accuracy than one collecting flat fees regardless.

"What if the operator games the system?" The same controls that prevent DARES gaming apply. Baselines are established during a measurement period. The IVA model — independent verification by a third party — can apply to software savings just as it applies to connection counts.

"Can a software company survive on performance fees?" Ask Omnidian, whose revenue tripled in two years. Ask the ESCO industry, which the IEA values in the tens of billions globally. If the platform genuinely delivers value, the vendor's share of verified savings can exceed what flat subscriptions would have earned. The difference is they have to earn it. Every month.

"What about free pilots?" This is where the model accelerates adoption. A 90-day pilot with no upfront cost removes every barrier to trial. After 90 days of real data showing real savings, the conversation shifts from "will this work?" to "how do we structure the ongoing relationship?" You are negotiating from demonstrated value, not projected ROI in a pitch deck.

Trust as a Business Model

There is a dimension to this that goes beyond spreadsheet economics.

Anyone who has done B2B sales in Nigeria knows that relationships and demonstrated value come before contracts. The operator burned by a vendor promising transformation and delivering a dashboard is not signing another annual contract on faith. The CFO who watched the naira halve while locked into a USD-denominated SaaS agreement is not enthusiastic about the next one.

Performance-based pricing is, at its core, a trust mechanism. It says: we believe in our technology enough to bet our revenue on your results. If we fail, we eat the cost. If we succeed, we share the upside.

Operators deploying sites with borrowed capital at 30% interest rates, collecting revenue in a depreciating currency, staring at a 12-month performance holdback that determines whether their project breaks even — they do not need another fixed cost delivering uncertain value. They need a partner structured the same way the grant programme is structured: outcome-first.

The World Bank's GPRBA exists because pay-for-outcomes drives better results than pay-for-activities. DARES embodies this. It is time for the software layer to catch up.

The Sector Needs a New Default

The mini-grid sector in Nigeria is entering its most critical phase. Over 900 sites are under construction. DARES is disbursing. Nobody is managing a fleet of 50 sites with WhatsApp and Excel — not if they want to hit their performance milestones, collect their grant holdbacks, and keep their diesel costs from eating their margins alive.

But the pricing model matters as much as the technology. A flat-fee SaaS subscription asks a cash-strapped operator to absorb risk the vendor should be sharing. A performance-based model distributes risk correctly: both parties profit only when outcomes improve.

The difference between a developer who collects their 12-month performance holdback on time and one who misses the deadline because their data was not in order can be tens of thousands of dollars per site. Multiply that across a fleet. Multiply that across the sector.

If the grant programme itself is structured around verified performance, why should the operational tools that enable that performance be priced any differently?

I would like to hear from operators, finance leads, and investors in this space. Does performance-based pricing for operational software make sense for your portfolio — or are there structural reasons it cannot work that I am not seeing?

It is 8:30 a.m. on the same Monday in Abuja. An operations manager at a mini-grid developer opens WhatsApp. There are 47 unread messages across six group chats. A site in Nasarawa has been running on diesel since Thursday because nobody logged the inverter fault. The technician who visited Saturday sent a blurry photo of the fault screen but no structured report. The SoC data from Victron VRM shows the batteries discharged to 18% before the genset kicked in, but SparkMeter Koios shows revenue collection was normal, so the issue went unnoticed for three days. The DARES quarterly report is due Friday, and the data has to be pulled from four systems and stitched together in Excel.

Both of these operators are competent professionals. The difference is not talent. It is tooling.

The Platforms That Developed Markets Take for Granted

In the US and Europe, a mature ecosystem of energy operations platforms has quietly become the operational backbone of fleets managing tens of gigawatts.

Stem / Athena (now PowerTrack). Stem built Athena, an AI-driven system that optimises battery storage dispatch in real time — deciding when to charge, discharge, and arbitrage based on grid pricing, weather, and demand. PowerTrack now monitors over 30 GW of solar assets; thirteen of the top fifteen US C&I operators standardise on it. The AI does not just monitor; it decides, re-evaluating market conditions every thirty minutes. Stem has struggled financially — a 69% revenue decline in 2024, a 27% workforce reduction — but the technology proved what AI-driven energy optimisation can do at scale.

Omnidian. If Stem is about optimising batteries, Omnidian is about guaranteeing solar performance. They guarantee that every asset under management will produce at least 95% of projected output, and if it falls short, they pay the difference. This works because their platform, Resolv, ingests system data around the clock and uses ML to detect degradation before it becomes a production loss. In 2024, Resolv detected nearly one million issues across 2 GW of managed assets. Omnidian recently raised $87 million; their revenue tripled between 2022 and 2024. This is what proactive O&M looks like when detection, dispatch, repair, and verification are integrated into a single platform.

Raptor Maps. Raptor Maps attacks the physical inspection layer using drones with thermal imaging and ML to detect panel-level defects across utility-scale and C&I portfolios. Their 2025 Global Solar Report drew on 193 GW of analysed capacity and found equipment-driven underperformance increased 214% over five years — an estimated $10 billion in unrealised revenue globally in 2024. They deploy autonomous drones at over 3 GW of capacity, and the entire workflow generates structured data that makes the next inspection smarter.

These three represent different layers of the same stack: optimise (Stem), guarantee (Omnidian), inspect (Raptor Maps). Together with Power Factors' Unity (18,000 sites) and AlsoEnergy's monitoring suite, they form a comprehensive ecosystem. The US operator's Monday morning is calm because these tools exist.

What Africa Actually Has Today

The African distributed energy sector is not starting from zero. Real tools exist.

Odyssey Energy Solutions acquired Ferntech to add remote monitoring to its platform, which serves as the operational backbone of Nigeria's DARES programme — 1,350 mini-grids, 17.5 million connections. Odyssey handles project tracking, grant disbursement, and basic monitoring. It is critical infrastructure. But it is primarily a programme management platform, not an AI-driven operations engine. Compliance reporting still requires manual data compilation.

AMMP Technologies is the closest analogue to what developed markets have. Monitoring nearly 1,000 systems across 17 countries, AMMP provides remote monitoring for everything from rural mini-grids to megawatt-scale PV-diesel hybrids. But with only $1.36 million raised, they are spread thin. No edge AI. No workforce management. No predictive maintenance. No Odyssey integration for DARES compliance.

SparkMeter Koios is essential for metering and revenue collection — the financial heartbeat of any mini-grid. Billing, payments, basic grid analytics, designed for low-bandwidth. But Koios is a metering platform, not an operations platform. It will tell you revenue was collected normally while your batteries are being destroyed by an undetected inverter fault.

Okra Solar's Harvest provides monitoring and billing for their mesh-grid hardware, but it is locked to Okra's ecosystem. If you are running Victron inverters with a FusionSolar array and SparkMeter metering, Harvest does not help you.

Each of these tools does something valuable. None of them does everything. And critically, none of them learns.

The Empty Half of the Matrix

When you map developed-market capabilities against what exists for African distributed energy, the top row — basic monitoring, revenue metering, remote visibility — is reasonably served. The bottom half of that matrix is almost entirely empty.

Predictive maintenance using ML. In the US, degradation patterns are caught weeks before failure. In African mini-grids, you find out when the system trips. The only operator running ML-driven predictive maintenance in Africa is Husk Power Systems — and it is proprietary, closed, and unavailable to anyone outside Husk's fleet.

AI diesel dispatch optimisation. For hybrid mini-grids running solar-battery-diesel, intelligent genset scheduling can reduce diesel consumption by 15-30%. No platform serving African operators offers this. Every litre wasted is naira burned.

Workforce and technician dispatch. In developed markets, work orders flow from detection to assignment to completion to verification in one system. In Nigeria, maintenance coordination happens on WhatsApp. No structured scheduling, no travel optimisation, no verification the repair worked.

Structured field data capture. Every repair in a developed-market fleet generates structured data: what failed, what was found, what was replaced, what performance looked like before and after. This data trains the models that predict the next failure. In African mini-grids, repair data lives in WhatsApp photos and phone calls. Nothing is captured. Nothing is learned.

Edge AI and offline operation. Rural sites frequently lose connectivity — harmattan dust on telecom towers, network congestion, no coverage. Any platform requiring constant cloud connectivity will have blind spots. Edge computing that processes locally and syncs when connected is standard in IoT. No platform serving African mini-grids offers it.

Fleet-wide learning. Perhaps the most consequential gap. When an MPPT controller fails during harmattan at a site in Kano, that failure teaches one company one lesson. No shared failure database. No cross-fleet anomaly detection. Every operator pays full tuition for lessons the sector has already learned.

Husk Power: Proof of Concept, Not a Solution

Husk Power Systems proves the concept works in Africa. Over 400 mini-grids across Nigeria and India. An integrated IoT and ML platform with demand forecasting, agentic AI optimising generation assets every thirty minutes, and predictive maintenance. EBITDA positive in Nigeria and India by Q4 2022. Now raising up to $400 million with a potential IPO in 2027.

But Husk's platform is proprietary. It is their competitive moat, and understandably so. No other operator in Nigeria — not developers deploying under DARES, not operators managing REA-funded sites — can access it. The proof of concept is locked inside one company.

The Opportunity Behind the Gap

Nigeria's REA has deployed over 1,000 mini-grids, with over 900 more under construction. DARES targets 1,350. Similar programmes are active across Kenya, Tanzania, Sierra Leone, and the DRC. The installed base is growing fast. The operational tooling is not keeping pace.

This is not a technology problem. Stem proved AI dispatch optimisation works. Omnidian proved ML-driven performance guarantees are commercially viable. Raptor Maps proved automated inspection generates compounding returns through structured data. These are solved problems.

The problem is that nobody has built these capabilities for African conditions: offline-first architectures for unreliable connectivity. Diesel-hybrid dispatch for systems that will run gensets for years. Integration with Victron VRM, FusionSolar, and SparkMeter — the actual hardware stack on Nigerian sites. Compliance automation for DARES and NEP reporting. Workforce management for technicians covering vast distances between rural sites. And pricing models that make sense for operators collecting revenue in naira.

The African energy sector does not need a watered-down version of what the US has. It needs purpose-built platforms designed for African conditions, informed by what the global market has already proven works. The bottom half of the capability matrix will not fill itself. Someone needs to build it.

Over 900 mini-grids are under construction in Nigeria right now. When they come online, how will they be operated — with WhatsApp and Excel, or with the kind of intelligent tooling that the rest of the world already relies on?

I would genuinely like to hear from operators, developers, and investors working in this space. What does your operations stack look like today? What is the biggest gap you feel every day?

But nobody can install them.

At the Nigeria Energy Conference in Lagos, Power Minister Adebayo Adelabu laid bare a number that should keep every energy sector leader awake at night: "If we bring all the six million meters that represent the gap in the sector today, it will take us five years to install them because we don't have installers."

Five years. Not because of funding. Not because of policy. Because of people.

This is the technician bottleneck. And if it's already crippling the grid metering program, imagine what it means for DARES.

The Math That Doesn't Work

Under the Distributed Access through Renewable Energy Scale-up (DARES) program, Nigeria is deploying 1,350 new solar mini-grids backed by $750 million from the World Bank, with over $1 billion in parallel private capital. These sites will bring clean electricity to 17.5 million Nigerians. It is, without exaggeration, the most ambitious rural electrification push in African history.

Each of those 1,350 mini-grids is a complex system: solar arrays, battery banks, inverters (Victron, Huawei, Growatt), smart meters (SparkMeter or similar), distribution infrastructure, and often a backup genset. Each one needs commissioning. Each one needs ongoing O&M. Each one will eventually have components that fail, degrade, or need recalibration.

Now consider the workforce available to maintain them.

NAPTIN's Director-General Ahmed Nagode has confirmed that Nigeria currently lacks 50% of the skilled manpower required to effectively run the power sector. Not 10%. Not 20%. Half. The country's most ambitious training response so far — the NextGen RESCO Programme, a joint initiative between REA, NAPTIN, and EIEI — secured about 130 paid renewable energy placements for young Nigerians in 2025. One hundred and thirty, against a deficit that numbers in the hundreds of thousands.

Zoom out further: the entire African continent accounts for only 344,000 renewable energy jobs, according to IRENA's 2025 Annual Review. That's for a continent of 1.4 billion people undergoing the fastest energy transition build-out in history.

The arithmetic is unforgiving. The government has announced plans to train 4,000 Nigerians in meter installation through NAPTIN. That's a good start. But DARES alone will need a sustained maintenance workforce across 1,350 geographically dispersed sites — many in locations where the nearest paved road is an optimistic description. Add grid-side metering, rooftop solar, interconnected mini-grids, and the broader NEP pipeline, and you're looking at a demand curve that training programs alone cannot bend fast enough.

What a Broken Visit Looks Like

To understand the bottleneck, you have to understand what a technician's day actually looks like in this sector.

Here is a scenario that anyone who has operated a fleet of mini-grids in Nigeria will recognize:

A community leader sends a WhatsApp message to the operator's regional office. "Site not working." The message gets relayed to a technician, sometimes within hours, sometimes the next day. The technician drives out — often 3 to 4 hours on secondary roads that, during harmattan or rainy season, can be barely passable. He arrives at the site with no diagnostic data. No fault codes. No equipment history. No idea whether the issue is the inverter, the battery bank, a tripped breaker, or a damaged distribution line. He troubleshoots on-site for an hour or two, determines it's a specific inverter board failure, but doesn't have the replacement part. He drives back. Sources the part. Returns days later. Another 3 to 4 hours each way.

That is 16 or more hours of skilled technician time consumed by a single fault — and the site was down the entire time, eroding community trust in solar with every dark evening.

In the field service industry globally, the best-performing companies achieve first-time fix rates of around 88%. Struggling companies sit at 63%. In the Nigerian mini-grid context — with poor diagnostic visibility, limited parts inventory systems, and vast distances between sites — first-visit resolution rates can fall well below that. Every repeat visit is a multiplier on an already scarce resource.

At 2 to 3 visits per incident, a technician covering sites spread across a state can realistically manage 8 to 12 sites. The rest of their time evaporates on the road, on callbacks, on guesswork.

Training Is Necessary. But It Is Not Sufficient.

Let me be clear: training more technicians is essential. Every NAPTIN cohort matters. Every NextGen RESCO graduate matters. The pipeline must grow.

But the gap is structural, not just numerical. Even if training programs scaled 10x overnight — and anyone who has worked in workforce development in Nigeria knows that scaling quality technical training is itself a multi-year infrastructure challenge — the deployment timeline for DARES cannot wait. Sites are being commissioned now. Communities are being connected now. Equipment is aging now.

The question isn't only "how do we get more technicians?" It's "how do we make every technician we have dramatically more productive?"

Every major industry that has faced a similar skilled-workforce bottleneck has eventually arrived at the same answer: technology that multiplies scarce human expertise.

The Pattern Is Clear

Healthcare figured this out. Nigeria has approximately 4 doctors per 10,000 people, far below WHO recommendations. The response wasn't solely to build more medical schools (though that's important too). It was telemedicine — connecting rural patients to specialists remotely, triaging digitally, ensuring that when a doctor does see a patient, the visit is maximally productive. Remote diagnostics, structured case notes, and AI-assisted triage mean each doctor's expertise reaches further.

Agriculture figured this out. U.S. farm output tripled between 1948 and 2021 while the farm workforce declined by 75%. Precision agriculture — GPS-guided equipment, soil sensors, drone monitoring, predictive analytics — didn't replace farmers. It made each farmer capable of managing vastly more land, with better outcomes. Yields increased 20 to 30% while input waste dropped 40 to 60%.

Telecom figured this out. Africa's tower companies manage hundreds of thousands of sites across remote terrain strikingly similar to where mini-grids operate. The shift to remote monitoring, IoT sensors, and predictive maintenance has dramatically reduced unnecessary site visits. When a field team does deploy, they arrive with exact fault data, the right parts, and a clear resolution path. The same number of technicians cover far more towers.

The pattern is consistent: when you can't scale the workforce fast enough, you scale what each worker can accomplish.

What This Looks Like for Mini-Grids

Imagine a different version of that broken-site scenario.

The inverter's operating parameters drift outside normal range at 2 AM. An intelligent monitoring system flags the anomaly, correlates it with the equipment's maintenance history and known failure modes for that hardware revision, and generates a structured work order before anyone wakes up.

By morning, the technician receives a notification — not a vague WhatsApp message, but a precise diagnosis: "Site 247, Victron MultiPlus-II, Error 11 — relay failure on L1, replacement board P/N VIC-2340. Estimated on-site time: 90 minutes." The work order includes GPS routing optimized for current road conditions, a parts checklist with barcode verification, and a step-by-step resolution guide.

The technician arrives prepared. One visit. Three hours on-site. Done. The site was down for hours instead of days. Community trust intact.

Now do the math in reverse. At one visit per incident with intelligent scheduling and route optimization, that same technician covers 20 to 30 sites instead of 8 to 12. You have effectively tripled your workforce without hiring a single additional person.

And every completed repair generates structured data that feeds back into the system — improving future diagnostics, refining parts forecasting, identifying fleet-wide patterns before they become fleet-wide failures. The system learns. Each technician's experience becomes institutional knowledge instead of disappearing when they move to a different employer.

The Race Against Degradation

Here is the uncomfortable truth that the DARES program will confront within its first 18 months of operation: solar assets degrade. Batteries cycle. Inverters have finite lifespans. Smart meters need firmware updates and occasional replacement. Distribution infrastructure in rural Nigeria takes a beating from weather, vegetation, and the occasional goat.

Without systems that predict failures, prioritize maintenance intelligently, and make scarce technician time maximally productive, the fleet will degrade faster than it can be maintained. We have seen this pattern before with donor-funded infrastructure across the continent — impressive commissioning ceremonies followed by a slow decline into dysfunction, not because the technology failed, but because the maintenance model was never viable at scale.

DARES is too important to follow that trajectory. The $750 million investment, the 17.5 million Nigerians waiting for reliable power, the climate commitments — all of it depends not just on building these mini-grids, but on keeping them running for 20+ years.

The workforce to maintain 1,350 mini-grids at the level they deserve will not materialize in time through training alone. The software layer — the intelligence that turns reactive maintenance into predictive maintenance, that turns blind site visits into precision interventions, that turns individual technician knowledge into fleet-wide learning — is not a nice-to-have. It is load-bearing infrastructure, as essential as the inverters and panels themselves.

The Question

We are building at a scale Nigeria's energy sector has never attempted before. The hardware is coming. The financing is in place. The communities are ready.

But are we building the systems that will keep these assets productive for the next two decades — or are we assuming a workforce will appear that the numbers tell us won't?

Arome is a CTO based in Abuja, building intelligent systems for the energy sector.

Sources

• Adelabu Decries Shortage of Skilled Power Technicians as 400,000 Meters Remain Uninstalled — Radarr Africa
• No Installers for 400,000 Imported Meters, Minister Flags Shortage — Lagos Television
• Nigeria Lacks 50% of Skilled Workforce Needed in Power Sector — Punch NG
• REA Secures 130 Energy Jobs for Nigerian Graduates Through NextGen RESCO Programme — Nigeria Education News
• Africa's Renewable Energy Jobs Hit 344,000 in 2024 — IRENA/ILO via Ecofin Agency
• Going Green: Scaling Up Access to Clean Electricity for Over 17 Million Nigerians — World Bank
• Nigeria to Expand Access to Clean Energy for 17.5 Million People — World Bank
• Benefits and Evolution of Precision Agriculture — USDA ARS
• IRENA Renewable Energy and Jobs: Annual Review 2025
• Nigeria's Road Network — Logistics Cluster Digital Logistics Capacity Assessments

Site 7 is a solar-hybrid mini-grid about four hours from your operations office. It was commissioned eighteen months ago, serves 340 connections, and has been performing well. But over the past three weeks, its lithium-ion battery bank has been quietly degrading — losing about 8% of usable capacity. Nobody notices. The operations manager responsible for that site is also responsible for 19 others. The degradation hides inside daily fluctuations of charge and discharge cycles, invisible in the spreadsheet she updates every Friday.

On Day 22, the battery fails. The genset kicks in at 100% runtime. Someone in the community sends a WhatsApp message. The technician gets dispatched, drives four hours, arrives without the right replacement parts because nobody knew exactly what had failed or how badly. He drives back. Procurement takes three days. He drives out again. Total elapsed time from first sign of degradation to resolution: 38 days. Total cost: somewhere between $3,200 and $7,800 when you add up the replacement battery, two round trips, excess diesel, and lost revenue.

This is not a hypothetical. This is Tuesday.

The Sector-Wide Warning Signs

Nigeria's mini-grid sector is at an inflection point. The Distributed Access through Renewable Energy Scale-up (DARES) programme, a $750 million initiative backed by the World Bank, aims to deliver clean power to 17.5 million Nigerians through solar mini-grids and standalone solar systems. More than 200 mini-grids were deployed in 2025 alone, and the pipeline is accelerating. Twenty-eight more were scheduled for completion in Q1 2026.

But deployment speed means nothing if the assets we are deploying cannot be sustained. NAPTIN (the National Power Training Institute of Nigeria) has warned that power sector reforms will stall without a skilled workforce capable of maintaining the infrastructure being built. The institute, in collaboration with JICA, has been developing training curricula for maintenance personnel — an acknowledgment that we are building faster than we can maintain.

And the evidence from the field confirms the concern. A Premium Times investigation that visited 40 mini-grid sites found systems "installed and commissioned without proper documentation — commissioning test results, as-built single line diagrams, key equipment manuals, maintenance spares, or operations and maintenance schedules." Dozens of government and donor-funded mini-grids were found operating far below capacity or lying abandoned entirely.

The pattern is consistent: we commission well, but we maintain poorly. And the cost of that gap is compounding.

What Reactive Maintenance Actually Costs

Let me walk through the economics that most fleet operators already feel but rarely quantify.

The multi-visit problem. In the current operating model, technician first-visit resolution sits around 40%. That means for every incident, there is a better-than-even chance your technician arrives, diagnoses the problem, and drives back for parts. At remote sites — and most mini-grid sites are remote by definition — each round trip costs $200-400 in fuel, vehicle wear, and staff time. Two to three visits per incident is the norm, not the exception.

The diesel trap. When a battery bank fails or an inverter goes down, the genset picks up the load. Diesel is already a dominant share of OPEX for hybrid systems — the SEforAll CAPEX/OPEX benchmark study found fuel can account for over 50% of ongoing costs in hybrid configurations. During an unplanned battery failure, diesel runtime goes to 100%. At current Nigerian diesel prices north of N1,800 per litre, an extended outage-response cycle can burn through $600-1,500 in excess fuel — compared to $100-200 if the disruption is caught early and managed intelligently.

The invisible revenue loss. Every hour a site runs on degraded capacity or is offline for repair is an hour of lost kWh delivered. For a site serving 300+ connections, that translates to $200-500 in lost tariff revenue per incident. Multiply that across a fleet of 20 or 50 sites, and the numbers become material to your P&L.

The compliance drag. For operators participating in results-based financing through NEP or DARES, performance data during outages must be manually compiled for IVA submissions and compliance reporting. When the data is fragmented across VRM exports, WhatsApp threads, and spreadsheets, compiling a single compliance report can take days of analyst time.

Add it all up, and a single undetected battery degradation event costs $3,200-7,800 by the time it is resolved. For a fleet of 20 sites experiencing even four such incidents per year, that is $256,000 to $624,000 in avoidable cost annually.

The Status Quo Tool Stack

Most operators I know are running some version of the same setup: Victron VRM (or FusionSolar, or SparkMeter Koios) for telemetry, WhatsApp for coordination, and Excel for tracking.

VRM is excellent hardware monitoring software. It was built for that purpose and it does it well. But it was not built to be a fleet operations platform. Its alert system will rate-limit emails if a site hovers near a warning threshold — a sensible anti-spam feature that also means critical alerts can go silent at exactly the wrong moment. Alerts arrive in your inbox alongside 47 other notifications, most of them false positives. The signal-to-noise ratio makes it nearly impossible to catch a slow degradation trend.

WhatsApp is fast, but it is not searchable, not structured, and not auditable. The diagnosis your technician texted about Site 7 last month? It is buried in a group chat between photos of someone's lunch and a forwarded news article.

Spreadsheets capture whatever someone remembers to enter. They do not capture what no one thought to look for.

The result: each failure produces zero structured training data. The lesson learned at Site 7 never transfers to Site 14 — even when the failure pattern is identical. The same $5,000 mistake repeats across your fleet because there is no system learning from it.

What Predictive Maintenance Makes Possible

Now consider the alternative.

Day 3: an anomaly detection system flags that Site 7's battery bank is showing an unusual capacity decline — not a threshold breach, but a trend that deviates from the battery's baseline behaviour. The pattern matches a known degradation signature.

Day 4: a work order is automatically generated. The system recommends specific replacement parts based on the diagnosed failure mode. The technician is dispatched with the right parts, the right tools, and a diagnosis already in hand.

Day 5: one visit. Three hours on site. Battery bank replaced. Site back to full solar operation.

Total cost: $800-2,200. Time from first anomaly to resolution: 8 days instead of 38. One technician visit instead of three. Minimal diesel burn. Full data capture for compliance reporting. And — critically — the failure pattern is now encoded in the system, so when Site 14's battery starts showing the same signature six months later, the detection happens on Day 1 instead of Day 3.

This is not science fiction. The U.S. Department of Energy has documented that functional predictive maintenance programmes deliver a 25-30% reduction in maintenance costs, 70-75% elimination of breakdowns, and 35-45% reduction in downtime. Companies like Omnidian are already monitoring over 1,500 MW of solar assets using machine learning to detect performance anomalies before they become failures.

The technology exists. The question is whether our sector will adopt it before the maintenance backlog becomes unmanageable.

The Learning Loop: Why Every Incident Should Make the Next One Cheaper

Here is what I think is the most underappreciated argument for predictive maintenance in the mini-grid context.

Nigeria is building hundreds of sites across dramatically different operating environments. A site in Borno faces different dust loading during harmattan than a site in Bayelsa dealing with humidity and salt air. Research on dust soiling in West Africa has shown that harmattan dust can reduce solar generation by more than 50% in some locations. Battery cycling patterns differ based on community load profiles. Inverter stress varies with grid interaction modes.

Every one of these sites is generating telemetry data. Every failure is a data point. But right now, that data evaporates. It sits in disconnected VRM instances, unstructured WhatsApp messages, and individual technicians' memories. When a technician leaves, his diagnostic intuition leaves with him.

A proper predictive maintenance system creates a learning loop: every incident feeds back into the model. Every resolved failure makes the next detection faster and the next dispatch more precise. Over time, the system learns your fleet — not just individual sites, but patterns across sites, across seasons, across equipment vintages.

This is especially powerful for the DARES pipeline. As hundreds of new sites come online, they will face the same failure modes that existing sites have already encountered. The operators who have been capturing and learning from their operational data will commission new sites with a maintenance playbook already written. The operators who have not will be starting from zero every time.

The Cost of Waiting

The Premium Times investigation found lead-acid batteries dying under Nigeria's heat, users disconnecting because of limited capacity, and communities losing trust in systems that were supposed to transform their access to electricity. Every abandoned mini-grid site is not just a financial loss — it is a setback for community trust that the next developer will have to overcome.

The SEforAll benchmark study recommended dedicating 20% of project budgets to maintenance. That is a start, but throwing more money at reactive maintenance is like buying a bigger bucket for a leaking roof. The question is not just how much we spend on maintenance, but whether we are spending it before the failure or after.

Reactive maintenance shortens asset lifespan by 30-40% and increases energy consumption by 15-20%. For a sector where the business case already depends on assets lasting 15-20 years, accelerating degradation through poor maintenance is an existential risk, not an operational inconvenience.

We are in a window right now where the fleet is growing fast but is still small enough to instrument properly. The operators who build predictive maintenance into their operating model today — who treat every site as a node in a learning network rather than an isolated installation — will be the ones still operating profitably in 2030. The ones who do not will be writing off assets at year five and wondering what went wrong.

The DARES programme represents the most ambitious bet Nigeria has ever placed on decentralized energy. The capital is flowing. The sites are being built. But capital expenditure without operational intelligence is just hardware waiting to become e-waste.

So here is my question for the mini-grid operators, developers, and funders reading this: if we know that reactive maintenance is the single biggest threat to mini-grid sustainability, why are we still treating predictive maintenance as a future investment rather than a present necessity?

I would genuinely like to hear how others in the sector are thinking about this.

Arome is a CTO based in Abuja building AI systems for the energy sector.

This person manages 18 sites. They are about to manage 45.

Welcome to the operational reality behind the largest distributed energy program the World Bank has ever undertaken.

The Scale Nobody Is Talking About

In December 2023, the World Bank approved the $750 million Distributed Access through Renewable Energy Scale-up (DARES) project for Nigeria — an IDA credit designed to provide over 17.5 million Nigerians with new or improved electricity access. The program became legally effective in late 2024, and the first grant agreements were signed in 2025.

The numbers are staggering. Up to 1,350 isolated mini-grids. Over 200 mini-grid developers expected to participate. A parallel deployment of 1.5 million solar home systems. $410 million allocated to mini-grids alone, with performance-based grants of up to $600 per verified connection.

This is not a pilot. This is industrial-scale deployment.

And it is happening in a country that installed 803 MW of new solar capacity in 2025 alone — a 141% year-on-year jump that made Nigeria Africa's second-largest solar market. Cumulative capacity hit 1.19 GW, with 96% of that being decentralized systems. Battery storage grew 305% in a single year.

The hardware is moving. The infrastructure is going into the ground. But here is the question nobody in the sector seems to be asking loudly enough: once these systems are commissioned, who actually runs them?

The Workforce Gap Is Not a Gap — It Is a Chasm

Let me put this bluntly. Nigeria lacks 50% of the skilled workforce needed to run its power sector. That is not my analysis — that is the Federal Government's own admission, delivered by the Minister of Power, who revealed that over 400,000 imported electricity meters are sitting in warehouses because there are not enough certified installers to put them on walls.

Four hundred thousand meters. Imported. Paid for. Gathering dust.

NAPTIN Director-General Ahmed Nagode has warned that many solar-powered systems, including streetlights in Abuja, could fail within five years without proper maintenance. He is not speculating. He is watching it happen in real time, in the nation's capital.

The entire African continent employs roughly 344,000 people in renewable energy, according to IRENA's latest data. That is 3% of the global total of 16.6 million. Nigeria's renewable energy sector employs approximately 70,000 people. We are deploying billions of dollars of infrastructure while the talent pipeline produces a trickle.

The REA, to its credit, launched the NextGen RESCO programme — an ambitious initiative that trained 130 young Nigerians in renewable energy skills. One hundred and thirty. From over 10,000 applicants. In a country that needs hundreds of thousands of skilled energy workers.

Training matters. But training alone cannot close a gap this wide, this fast.

What Operations Actually Look Like at 15 Sites

To understand why this matters, you need to understand what day-to-day mini-grid operations actually look like in Nigeria today.

A typical developer running 15 sites across two or three states is managing a patchwork of disconnected systems. Victron VRM handles the inverter and battery monitoring for Victron-equipped sites. FusionSolar covers the Huawei installations. SparkMeter Koios manages prepaid metering and revenue collection. WhatsApp is the de facto dispatch system for technician coordination. Excel does everything else — compliance tracking, diesel runtime logs, tariff cost build-ups, grant disbursement milestones.

Five systems. Five logins. Five data silos. Zero integration.

When a site goes down at 2 AM, someone notices the SoC dropping on VRM — if they happen to be checking. They send a WhatsApp message to the nearest technician. The technician drives out, diagnoses the issue, reports back via voice note. Someone updates the spreadsheet. Maybe.

At 15 sites, this works — barely — through heroic individual effort. The kind of effort where one operations manager knows every site by heart, recognizes anomalies by gut feel, and personally follows up on every unresolved ticket because there is no ticketing system.

But DARES is not 15 sites. For a mid-sized developer participating in the programme, we are talking 50 to 150 sites. At that scale, the WhatsApp-and-spreadsheet model does not degrade gracefully. It collapses.

The Performance Trap

Here is where the operational gap becomes a financial crisis.

Under the DARES performance-based grant structure, developers receive up to $600 per verified connection, but the money does not flow all at once. Forty percent is disbursed after commissioning. The remainder is tied to sustained performance, with connections verified by an Independent Verification Agent (IVA) through the Odyssey platform, phone calls, and field inspections.

This means a developer with 5,000 connections across multiple sites has hundreds of thousands of dollars locked behind their ability to prove ongoing service delivery. Monthly updates on Odyssey are not optional — they are the mechanism through which grant disbursement happens.

Now layer on the economics. According to regional benchmarking data, plant O&M costs represent roughly 33% of the total tariff cost build-up for mini-grids. Fuel accounts for 51% of operating expenditure. These are not margins that forgive operational sloppiness. A site running its genset two hours longer than necessary because nobody caught a PV yield drop is burning cash. A meter fault that goes undetected for a week is lost revenue. A missed IVA deadline is deferred capital.

When your grant disbursement, your revenue collection, and your operational costs all depend on real-time awareness of what is happening across your fleet — and your primary management tool is a WhatsApp group — you have a structural problem.

The Math That Training Cannot Solve

Let me run some numbers.

If 200 developers deploy an average of 7 sites each in the first phase, that is 1,400 sites requiring daily monitoring, regular maintenance, and continuous performance reporting. Each site needs routine O&M visits, remote monitoring, meter management, community engagement, and compliance documentation.

A well-run operation can perhaps manage 10-15 sites per operations team with current tools. At DARES scale, the sector would need somewhere between 90 and 140 dedicated operations teams — call it 500 to 700 skilled operations professionals, minimum, just for the mini-grid component. This does not include the 1.5 million solar home systems.

Where are these people? Nigeria trained 130 through NextGen RESCO. Even if every technical training programme in the country — REA, NAPTIN, GIZ, USAID, private academies — collectively produces 500 graduates per year, the pipeline cannot fill the demand created by a $750 million deployment wave.

This is not an argument against training. Training is essential. But it is an argument that the sector needs to think differently about how operations are structured. If you cannot hire your way out of the shortage, you need to multiply the effectiveness of every person you do have.

The Real Question

The Nigerian mini-grid sector has spent the last five years solving the financing problem. DARES represents that problem being solved — at scale. Performance-based grants. Results-based financing. IDA credit. Parallel financing from JICA, USAID, GIZ, AfDB.

The next five years will be about solving the operations problem.

The sector is about to deploy $750 million worth of distributed energy infrastructure. The grant structure demands sustained performance. The workforce does not exist at the scale required. The tools most operators use today were not designed for fleet management — they were designed for individual site monitoring.

Something has to give. Either the operational model evolves, or the performance metrics slip, the grant holdbacks accumulate, and the programme underdelivers on its promise to 17.5 million Nigerians.

I do not think that is inevitable. But I think the sector needs to be honest about where we are. We have solved the "how do we fund this" question. We have not solved the "how do we run this" question.

The hardware is going into the ground. The grant agreements are being signed. The clock is ticking.

So here is what I want to ask this community: if you are a developer scaling from 10 sites to 50 under DARES, what does your operations plan actually look like? What tools are you using, what is breaking, and what would you build if you could start from scratch?

Sources referenced in this article:

• World Bank approves $750M DARES project for Nigeria (December 2023)
• Nigeria lacks 50% of skilled power sector workforce — The Punch
• 400,000 imported meters uninstalled due to installer shortage — Radarr Africa
• NAPTIN DG warns solar systems could fail within 5 years — The Whistler
• REA NextGen RESCO trains 130 young Nigerians — ThisDay
• Africa has 344,000 renewable energy jobs — IRENA/ILO 2025 Review via Ecofin Agency
• Nigeria adds 803 MW solar in 2025, 96% decentralized — Ecofin Agency
• Nigeria's renewable sector employs ~70,000 — Leadership
• Odyssey as digital backbone for DARES — Odyssey Energy Solutions
• Mini-grid CAPEX and OPEX benchmarks — SEforALL, August 2024
• RMI on closing Nigeria's green skills gap — RMI