For decades, the power grid was calm and predictable, and, importantly, electricity flowed one way, from power plants to homes. Utilities managed it like clockwork.

But today’s grid is a different story — faster, more dynamic, and a lot less predictable. From sudden power fluctuations to intermittent outages and unusual load patterns, abnormal conditions can now appear anywhere and anytime. If left unchecked, these grid “gremlins” can swiftly spread throughout the network and disrupt operations.

What’s Behind the New Complexity?

Several factors give rise to the challenges of managing today’s grid. First, we’re using more electricity than ever before — charging electric vehicles, heating homes with electric pumps, running energy-hungry smart devices. Add to that growing demand complexity introduced by producing and storing power locally, through rooftop solar, batteries, and even EVs, and allowing the power from these Distributed Energy Resources to be fed back into the grid.

While this can be great for flexibility and sustainability, it also transforms the grid from a one-way highway into a complex, two-way energy ecosystem where conditions are constantly shifting, making it harder to manage. In this environment, a single surge in EV charging or a sudden dip in solar production can ripple across the network, creating reliability issues and wreaking havoc on operations.

The Role of AI and Edge Computing

To manage this new complexity, utilities are turning to AI-powered analytics and edge computing. Here’s how Landis+Gyr is putting this powerful combination to work for utilities.

Real-Time Intelligence at the Edge

Because managing a dynamic, two-way grid requires more than just visibility, utilities need solutions that embed real-time intelligence directly into the grid itself — inside smart meters, transformers, and substations — instead of relying only on centralized systems. This real-time, localized intelligence empowers the grid to:

1. Detect and diagnose anomalies instantly
   
2. Predict load spikes and distributed energy swings
3. Automatically balance local supply and demand
4. Take immediate action — without waiting for instructions from a central hub

In other words, every part of the grid can now sense, think, and act for itself.

At the heart of this solution is Revelo®, Landis+Gyr’s advanced IoT-enabled smart meter. Revelo captures high-resolution waveform data — offering deep, granular insight into everything from voltage quality to load behavior and asset performance. With the data sensing capabilities of Revelo, utilities can:

1. Identify abnormal patterns or "gremlins" such as voltage sags, surges, or equipment malfunctions
2. Predict potential faults before they cause outages
3. Trigger alerts and automated responses to mitigate risks

Harnessing AI: Connecting Edge and Cloud for End-to-End Insights

While intelligent edge devices handle immediate detection and response, aggregated data is also sent to Landis+Gyr’s cloud-based analytics platform. In the cloud, advanced AI and machine learning take over to:

1. Validate critical models, including meter-to-transformer mapping and phase identification 
2. Monitor and maintain high standards of power quality
3. Detect patterns and identify assets, with capabilities like pattern detection, capacity contribution analysis, revenue protection, and EV detection
4. Assess grid impacts and explore potential solutions

This edge-to-cloud integration ensures that utilities not only react faster — they plan smarter.

Staying Ahead of Disruption – and Gremlins

As grid demands grow more complex, so does the risk of service interruptions. By transforming everyday grid devices into intelligent agents, Landis+Gyr is helping utilities stay ahead of disruption, reduce downtime, and deliver more reliable, efficient service to their customers. With Revelo’s intelligent edge computing capabilities and high resolution data capture, combined with AI-based advanced analytics, utilities can build a grid that’s no longer reactive — it’s responsive, adaptive, and future-ready.