The artificial intelligence (AI) boom is poised to fuel a rapid and exponential surge in electricity demand, putting unprecedented pressure on utilities to modernize the grid, integrate distributed energy resources, and reduce increasing supply chain and customer costs.
Perhaps ironically, one way to address AI-driven power demands is by deploying AI tools. AI platforms have the potential to monitor faults in the grid, predict periods of high demand, and help systems run more efficiently.
This is the second part of a three-part series on the impact of artificial intelligence (AI) on utility operations. You can read part 1 here. Learn more about AI and power generation below. powers Here we continue to cover AI and its impact on the power sector.
But as IBM’s energy industry GM puts it, utilities “prefer to be quick to catch up” when it comes to implementing new technology, because there are significant risks for an organization whose mission is literally to keep the lights on.
By understanding how other sectors are beginning to leverage AI today, utilities can identify proven applications, accelerate adoption, and begin the transition from experimentation to value creation.
Use cases from other industries and how utilities can apply them
Utilities can learn valuable lessons about AI adoption by considering applications in a variety of industries, from automotive and manufacturing to firefighters and retailers. Here are some important use cases to keep in mind.
Visual Anomaly Detection – Insights from the Automotive Industry: Automotive manufacturing is using AI-driven computer vision to inspect parts and assemblies while also detecting defects such as misaligned welds, paint scratches, and missing parts. For example, major automotive OEMs like Toyota are applying AI to the paint inspection process and final vehicle inspection.
How utilities can apply it:
- Deploy high-resolution cameras, drones, or video feeds along power lines, substations, and remote infrastructure. Live image/video data is then fed into an AI model, which is trained to detect anomalies such as bent hardware, loose fittings, vegetation encroachment, and corrosion.
- Real-time alerts when anomalies are detected allow for early intervention rather than waiting for scheduled inspections or failures.
- Combining visual inspection with sensor and environmental data to prioritize field workers, reduce downtime, and extend asset life.
Wildfire monitoring – borrowed from high-risk landscapes: In fire-prone western states, governments and third-party providers are beginning to leverage imagery, drones, and AI to monitor wildfire ignition risk.
How utilities can apply it:
- Using AI models, it continuously scans video or still image feeds of assets in hallways, remote lines, and high-risk zones for signs of risk such as thermal anomalies, smoke, sagging lines, and damaged structures.
- Automate your decision chain. Trigger inspections immediately based on exceptions when risk thresholds are exceeded, rather than waiting for a planned physical system review.
- Combine internal maintenance data with weather, vegetation growth, and asset data to generate predictive risk maps that help you allocate resources more effectively.
Enhancing customer service and system design optimization – Lessons from retail and telecom: Industries such as retail and telecom are using AI bots and algorithms to improve customer experiences and design processes. For example, in system design, AI tools can be used to suggest system layouts that optimize cost, accuracy, schedule time, and potentially use less material.
How utilities can apply it:
- Deploy conversational AI systems (chatbots/virtual agents) to handle routine customer inquiries such as billing, service requests, and outage reports, freeing up human staff to focus on complex issues.
- Use design automation AI in system planning. For example, when designing grid extensions, AI can generate alternative layouts, cost-performance tradeoffs, and material requirements.
- Employ generative design programs to simulate “what-if” scenarios in system architectures, allowing planners to evaluate more options faster.
Supply Chain, Inventory, and Materials Tracking – Manufacturer’s Handbook Page: Manufacturing companies use AI to track inventory in real-time, predict parts demand, identify supply chain bottlenecks, and optimize inventory levels.
How utilities can apply it:
- Monitor critical spare parts inventory using AI. Forecast demand based on outage history, asset age, weather cycles, and project timelines. Automatically trigger replenishment and supplier sourcing.
- Implement AI to evaluate multiple suppliers, lead times, delivery risks, and cost options in near real-time.
- Reduce overstocking and understocking by integrating inventory data, project schedules, and asset condition monitoring.
Data-driven decision support – from static dashboards to prescriptive analytics for Industry 4.0: Utilities have already invested heavily in sensors, smart meters, grid software, and dashboards over the past decade. The challenge now is to act on that data, like the best of Industry 4.0, and use it to improve decision-making.
How utilities can apply it:
- Move beyond static dashboards to AI-driven analytics that recommend next steps or automatically trigger next-best actions.
- Use AI to analyze historical operational data and real-time sensor streams to predict and proactively intervene in equipment failures, system stress, load imbalances, or peaks in customer demand.
- Identify operational inefficiencies that might be missed by human analysis, reducing costs and improving reliability.
overcome the fear factor
Deploying AI across utilities is not a passing fad. As the demand for electricity soars, electricity is rapidly becoming a strategic necessity.
However, it is natural for power companies to be hesitant. They run critical infrastructure and collect sensitive customer data. This increases the risk of introducing new technologies, especially new tools such as AI.
Expect bumps along the road. After all, utilities are creating something new. However, learning from other industries can be an efficient and exciting first step in charting a path to future success.
—andrew bourdain He is the head of Actalent’s Grid Automation practice.
