As energy systems become more distributed, digital and dynamic, utilities are increasingly turning to AI and IoT to help modernize operations, improve reliability and accelerate decarbonization.

Yet according to the IDC InfoBrief, How AIoT Is Reshaping Industrial Efficiency, Security, and Decision-Making, more than half of AIoT initiatives are still stuck in pilots or limited rollouts. What’s holding organizations back – and what differentiates those that scale successfully?

To explore these questions, we sat down with Adam Sroka, CEO and Co-Founder of Hypercube, whose team works closely with global energy companies on real-time optimization, forecasting and AI-driven operational decisioning. In this Q&A, Sroka shares where AIoT is delivering the strongest value today, how organizations are closing the workforce skills gap, and the capabilities he expects to define the next two years.

IDC found that 57% of AIoT deployments are still in pilot or limited rollout.

Q: From your perspective, working with energy companies, why is AIoT adoption still stuck in early stages, and what differentiates the utilities that successfully move from pilots to widespread deployment?

Adam Sroka: AIoT adoption is slow because many organizations don’t yet have the infrastructure needed for real-time, data-driven operations. Devices are connected, but the data they produce is often inconsistent, siloed or not ready for analytics and automated decision-making. Without strong data pipelines and clear governance, modern cloud architecture, pilots remain isolated rather than scalable.

The organizations that progress invest early in these foundations and choose AIoT projects with clear operational or financial outcomes. They also establish shared ownership across data, IT and OT so AI becomes part of core operations rather than an experimental add-on. When these conditions are met, moving from pilot to widespread deployment becomes predictable rather than aspirational.

AIoT significantly enhances operations, boosting security, increasing revenue, improving compliance and streamlining workflows (32–37%).

Q: Where does Hypercube see AIoT driving the biggest operational improvements for energy companies – grid reliability, field operations, asset performance, regulatory compliance, or something else?

Sroka: AIoT delivers the most impact when organisations use real-time device data to actively guide operational decisions. For grid and portfolio operators, the greatest benefits arise from enhanced visibility and predictability. As networks integrate more distributed and flexible assets, reliable insight into how these assets behave becomes essential. AIoT enables operators to identify developing issues earlier, understand localized interactions across the system and respond with more confidence.

The same data-driven approach enhances asset performance by identifying inefficiencies, degradation patterns, and opportunities to optimize equipment usage throughout the day. Aggreko is a good example of this, where automating IoT-driven monitoring and alarm processes has led to faster and more consistent operational oversight of critical power assets.

AIoT also enhances day-to-day operational efficiency by reducing the manual work required to maintain and monitor complex systems. Automated analysis and alerting allow teams to focus on higher-value tasks rather than routine checks, and responses become more consistent because they are based on structured logic rather than ad hoc judgment.

As regulatory expectations for data transparency and system assurance increase, AIoT provides clear audit trails, traceable data flows and predictable model behavior. Taken together, these improvements enable energy companies to run more reliable, efficient and compliant operations at a time when the system is becoming significantly more dynamic and demanding.

Skill shortages are now the #1 barrier to AIoT adoption – up from #5 in 2019.

Q: IDC highlights a skills gap as the biggest challenge to scaling AIoT. What strategies are working for energy organizations to upskill teams, especially in OT environments where AI expertise is scarce?

Sroka: The skills gap is widely recognised, but the most effective organizations address it by designing AIoT systems that reduce reliance on deep technical expertise in day-to-day operations. Instead of expecting engineers to understand machine learning models, they focus on building predictable data pipelines, consistent deployment processes and interfaces that present clear outputs. This reduces the cognitive load on operational staff, allowing teams to work confidently with AI-generated insights without needing to build or maintain the underlying models themselves.

Upskilling still matters, but it works best when it is targeted and tied directly to real workflows. Successful energy organizations provide training that helps teams interpret forecasts, understand anomaly alerts, and apply optimization recommendations in context, rather than delivering generic AI theory. Many organizations also strengthen their capability by working alongside specialist consultancies that can provide the breadth of skills and energy-sector experience that would be difficult to develop or hire internally.

This model allows organizations to move faster without carrying permanent overhead, while their teams learn through direct collaboration and structured knowledge transfer. The combination of simpler systems, targeted upskilling and access to specialist expertise creates a realistic and sustainable way to close the skills gap.

Organizations that heavily utilize AIoT are twice as likely to report benefits that significantly exceed expectations. Less than 3% say AIoT fell short.

Q: What differentiates the energy companies that see AIoT delivering “significantly above expectations”? Are there specific practices, governance models, or technology choices that correlate with higher value?

Sroka: High-performing organizations treat AIoT as a structured operational capability, not a collection of experiments. They focus on well-defined use cases, establish governance early and ensure that models and data flows are understood by the people who rely on them. In our work with Heatio, this approach was central to building a real-time optimization platform capable of ingesting data from domestic batteries, heat pumps, solar systems and smart meters, then using that information to make reliable, automated decisions for customers.

Another differentiator is that these organizations also invest in platforms that can support multiple future use cases rather than building one-offs. When IT, data and operational teams share ownership of these platforms and their governance, AIoT becomes scalable and dependable and the value compounds as more device types and optimisation capabilities are added.

64% of organizations expect moderate to significant AIoT growth in the next two years.

Q: Based on IDC’s findings and your experience in the field, what do you think will be the defining AIoT capabilities energy companies adopt in the next 12–24 months?

Sroka: In the next two years, forecasting is likely to become more closely integrated with operational systems. Instead of being used mainly for analysis or reporting, we expect to see AI forecasts play a more active role in informing scheduling, dispatch and trading decisions as organizations gain confidence in the underlying models. Organizations will also adopt more supervised automation, where AI recommends or executes decisions under defined guardrails.

Monitoring and assurance will become part of standard practice. Companies will expect continuous insight into model performance, data quality and system behaviour to ensure reliability and compliance as more processes become automated. These shifts will move AIoT from a peripheral technology to an integral part of everyday operations.

Q: If you could give one piece of advice to an energy executive looking to accelerate AIoT investments, what would it be?

Sroka: Energy executives should prioritise building a strong data and cloud foundation and delivering one or two AIoT use cases end-to-end before attempting broad expansion. Proving clear value early creates internal confidence and establishes a repeatable delivery pattern.

This approach ensures that AIoT becomes a reliable operational capability, rather than a series of disconnected pilots and positions the organization to scale efficiently once the first successes are in place.