AI and smart meters transform African utilities

Utilities across Africa have entered a new era of digital grid management, with AI, smart metering and real-time operational visibility emerging as critical tools in managing electricity networks.

In a panel discussion at Enlit Africa in Cape Town , participants outlined how data-driven technologies are reshaping grid operations, improving forecasting accuracy and strengthening system resilience across the continent.

Moderated by Pitso Sekhoto, Middle Manager at the National Transmission Company of South Africa (NTCSA), the panellists explored the operational realities of deploying advanced metering infrastructure (AMI), managing massive data streams and leveraging AI to transform raw data into actionable insights.

Protecting the grid

Herman Mare, general manager for protection and control at ACTOM, traced the evolution of grid protection systems from electromechanical relays to highly digitised, interconnected networks.

“Electromechanical relays were similar to old electromechanical metres. You essentially only knew whether there was power or no power, or whether a trip had occurred. Digitisation has fundamentally changed the amount of information available from protection systems,” said Mare.

Modern utilities can now monitor networks on a national scale in real time, with data flowing in milliseconds, he added.

This enables operators to design wide-area protection systems capable of responding to events occurring hundreds of kilometres apart.

“Something happening in Mpumalanga can now influence a protection decision in Cape Town,” he said.

Mare described the development as “revolutionary” for grid protection and system safeguards.

But he cautioned that digitalisation also introduces new operational risks, particularly around skills shortages.

Traditional protection engineers, he noted, were trained to work with physical tools such as multimeters and screwdrivers, whereas modern grids require expertise in data analytics, network communications and fibre optic systems.

“We now need data analysts and people who understand layered network communications. The challenge is how to transition technicians with 30 or 40 years of field experience into maintaining digital networks.”

While cybersecurity remains a concern, Mare argued that the greater immediate risk for utilities is the widening skills gap required to operate increasingly digital infrastructures.

The ESG elephant in the room

The growing importance of data visibility was echoed by Zama Mkhize Pila, group ESG manager at Dis-Chem, who highlighted the connection between electricity consumption, emissions reporting and corporate sustainability strategies.

“At the centre of any economy is electricity. The link between ESG and electricity use has not really been fully made.”

She highlighted that businesses are increasingly collecting detailed electricity consumption data through metering and monitoring systems.

This enables organisations to track emissions, improve operational efficiency and support decarbonisation strategies.

“Data helps us identify inefficiencies and engage more intentionally with operations. Boards now want detailed insight into energy use, carbon accounting and future sustainability targets,” said Pila.

AI and advanced analytics would become increasingly important in helping companies move from analysing historical data towards predictive planning and proactive energy management.

“There’s a shift from looking backwards at historic data to asking what this data is telling us about the future,” she said.

For utilities, she argued, corporate energy data could also support broader national planning efforts, particularly as renewable energy deployment accelerates across South Africa and the continent.

Have you read?
Interoperability critical to long-term smart grid investment
Powering utilities via smart metering and energy intelligence

Data provides visibility for utilities

The importance of visibility at network level was reinforced by Cleophas Ogutu, project manager for GIS (Geographic Information Systems) data clean-up and loss reduction at Kenya Power.

Ogutu described accurate network visibility as the “foundation” for reducing both technical and non-technical losses in utilities.

“You need to know where your assets are, where your metres are connected and how they are mapped to the grid. Without visibility, utilities are effectively walking blind,” he said.

Referencing Kenya Power’s digitalisation journey, Ogutu said improved GIS visibility had enabled the utility to separate technical losses from commercial losses and allocate operational resources more effectively.

He warned, however, that poor quality GIS data remains one of the greatest risks facing utilities attempting to deploy AI forecasting and smart metering systems.

“All these major investments… depend on clean, structured and complete data. If they are built on poor data, the investment becomes useless.”

Why AI is becoming more critical for utilities

Al’Louise Van Deventer, general manager for engineering and technology at South African utility Eskom, described data as “the eyes and ears of the operator”.

“Without visibility, operators cannot make the decisions needed in real time to protect the grid,” she said.

Van Deventer pointed to global blackout events as evidence that failures in operational visibility and data management can have severe consequences.

As more distributed technologies such as rooftop solar systems and electric vehicle (EV) charging stations connect to networks, utilities require increasingly detailed real-time visibility to maintain safety and grid stability.

“If a line is isolated but operators don’t have visibility that it remains live because of embedded generation, people can be electrocuted,” she said.

Van Deventer also highlighted the enormous data volumes now being generated by smart meters, noting that Eskom has already deployed more than two million smart meters, creating continuous streams of operational data.

“The data never stops. You cannot simply switch it off.”

Managing these datasets, she argued, has made AI “almost critical” for utilities.

AI systems are increasingly needed not only to process data volumes but also to prioritise meaningful insights while filtering out operational “noise”.

“You cannot reach advanced levels of modernisation without AI. It becomes impossible to process those data sets manually.”

While African utilities face significant challenges around data quality, skills development and infrastructure modernisation, digitalisation and AI are rapidly becoming essential tools in building more resilient, efficient and customer-focused electricity systems, the panel concurred.