The global energy system is entering one of the most complex transitions in its history. Utilities are being pulled in two seemingly opposite directions at once: electricity demand is surging, fueled by electrification, digitalisation, and especially the explosive growth of AI-powered data centers, while pressure mounts to decarbonise faster than ever before.
Itron’s Stefan Zschiegner brings focus to a striking paradox. Artificial intelligence is both driving energy consumption and emerging as one of the most powerful tools to manage it. Far from creating a stalemate, this contradiction is becoming a powerful engine of innovation.
The numbers underscore the urgency. Morgan Stanley estimates that by 2030, data centers alone could consume up to 8% of total national electricity demand in countries such as Australia. At the same time, electricity networks are being reshaped by the rapid spread of electric vehicles, rooftop solar systems, and residential battery storage. What used to be a simple one-way flow of power, from centralized generators to passive customers, is evolving into a highly decentralized, two-way network of millions of active participants. Traditional grid models built on centralized command-and-control architectures are no longer capable of managing this complexity at scale.
Distributed intelligence: a complementary evolution to conventional AI
While traditional AI tends to concentrate data processing and decision-making within centralized computing facilities, distributed intelligence disperses those capabilities across the grid itself. Analysis and control shift outward to the network edge, where data is generated and action must occur in real time.
This change turns once-simple measuring devices into active grid managers. Smart meters equipped with distributed intelligence act not just as sensors, but as controllers, capable of assessing local grid conditions and autonomously dispatching distributed energy resources (DERs), including EV chargers and solar inverters, without waiting for instructions from a central operations center.
The benefits of this architecture are immediate and tangible. Latency drops. Grid resilience improves. Local optimization becomes continuous rather than reactive. In real-world applications, distributed intelligence-enabled meters can recognize EV charging patterns, detect meter bypass activity, and predict transformer overload risks in real time, coordinating locally while remaining linked to centralized control systems.
DERs: From Grid Disruption to Grid Advantage
Solar installations, residential batteries, EV chargers, and smart appliances are no longer fringe technologies. They now represent the foundation of a flexible, decarbonised energy system. But without intelligent orchestration, these resources risk adding volatility rather than stability to grid operations.
Utilities therefore require control platforms capable of managing millions of edge devices simultaneously, blending decentralised AI capabilities with real-time forecasting, localized automation, and DERMS (Distributed Energy Resource Management Systems).
At scale, orchestration systems must support:
- Flexible, least-cost integration of diverse DER portfolios
- Live load shaping and grid constraint management
- Market participation via open standards, including IEEE 2030.5, OpenADR, and similar protocols
- Transformer protection and outage awareness powered by distributed intelligence
The challenge is not whether DERs can support grid transformation, it is how effectively they can be coordinated across an increasingly decentralized system.
Early deployments of distributed intelligence and DER integration have gained momentum in North America and Australia, but the model is universally relevant. The same principles apply whether utilities are managing EV charging peaks in Norway, rooftop solar saturation in California, or community microgrids in Southeast Asia.
Across geographies, utilities face a common set of pressures: aging infrastructure, changing regulation, rising reliability expectations, and accelerating decarbonisation mandates. Distributed intelligence offers a practical way forward, a solution that is scalable across networks, adaptable to local grid conditions, and secure enough for mission-critical operations.
When AI-driven distributed intelligence converges with DER deployment, utilities unlock a wide spectrum of benefits:
- Real-time consumption insights, proactive alerts, and personalized energy services create deeper customer interactions and stronger trust.
- Localized autonomous control improves outage response, manages volatility, and supports decarbonisation goals through continuous balancing.
- Utilities gain the ability to build virtual power plants, activate DER marketplaces, and enable wholesale participation across ISO, RTO, and TSO balancing authorities, supporting regulatory frameworks such as FERC Order 2222 in North America and NEM reform in Australia.
- Reduced truck dispatches, faster restoration times, and predictive maintenance translate to lower operating costs and more responsive field operations.
While many deployments begin with limited use cases such as managing minimum demand at a small population of endpoints, the real objective is scale. Systems must ultimately coordinate millions of DERs and enable sophisticated applications, including ancillary grid services and detailed load-curve shaping that stabilize distribution networks. AI at the edge becomes essential for managing this growing complexity.
The transformation of grid devices is profound. DERs are no longer passive assets sitting at the network fringes. Through distributed intelligence, each connection point becomes a decision-making node, capable of sensing, predicting, and responding in real time. What was once merely an “endpoint” evolves into a grid “decision point.”
Achieving a sustainable energy future requires far more than simply adding renewables to the generation mix. It demands intelligent orchestration across every layer of the grid. By embedding AI at the network edge through distributed intelligence and integrating DERs into adaptive control systems, utilities can build electricity networks that are more resilient, more efficient, and fundamentally smarter, while advancing decarbonisation goals.
