The improvement cycle in traditional AI can run months long, requires thousands of labeled examples, and still doesn’t guarantee results. For industries like energy infrastructure, where the most critical defects are also the rarest, that’s not a minor inconvenience. It’s a dead end.

So we built our own patent-pending AI to solve it.

This whitepaper covers our Adaptive AI approach – which flips the model entirely: instead of collecting data, labeling it, training for weeks, and hoping the output reflects what your experts actually meant, domain experts simply show the system a few examples. It learns. It adapts in real time. And when better foundation models come out, performance improves automatically: no retraining, no re-labeling, no waiting.

Built on eSmart Systems’ patent-pending Adaptive AI, the new platform lets utilities and technology companies build, deploy, and scale custom AI models in minutes — without machine learning expertise or retraining cycles, directly accessible for AI Agents.

March 12, 2026 — eSmart Systems, the market leader in computer vision for electric utility infrastructure, today announced the public launch of AI Studio by eSmart Systems: a full-stack AI development and deployment platform that makes the company’s proprietary computer vision intelligence available as modular, API-first services for the first time.

AI Studio enables utility teams, technology companies, and inspection professionals to build, tune, and deploy custom AI models aligned to their own operational standards and asset definitions — without requiring an in-house machine learning team.

At the core of AI Studio is eSmart Systems’ patent-pending Adaptive AI technology: a methodology that enhances performance from few-shot learning to generate working detectors and classifiers from a small number of images in minutes, adapting in real time to new definitions and edge cases without traditional retraining cycles. The result is up to 35 percentage points accuracy ahead of competing few-shot technologies and ~99% accuracy across every benchmark, with zero retraining time.

Solving a Decade-Long Barrier to Utility AI Adoption

For utilities, the promise of AI-powered grid inspection has long been limited by a fundamental challenge: generic computer vision models are not built for the diversity of real-world infrastructure assets, local inspection standards, or the operational requirements of transmission, distribution, and substation workflows. Customizing AI to meet those standards has historically required significant internal machine learning expertise — a resource most utilities and inspection teams do not have.

AI Studio removes that barrier entirely.

Three Capabilities Powering an End-to-end Model Development Platform

AI Studio is built around three core capabilities, accessible through both a web-hosted interface at ai.esmartsystems.com and fully documented APIs:

• Model Builder: Create new custom models from a handful of example images using Adaptive AI. No annotation pipelines, no training runs, no ML expertise required. Models continuously adapt in real time as definitions evolve or edge cases emerge.
• Model Garden: Access to eSmart Systems’ library of production-ready detectors, classifiers, and pipelines, built on millions of power utility images across transmission, distribution, and substation domains. Continuously refined from real-world deployment across 75+ utilities globally.
• Pipeline Builder: Compose multi-step vision workflows visually using a drag-and-drop editor — for example, detection → classification → measurement — with instant API deployment and full input/output transparency for audit and compliance purposes.

Every model and pipeline in AI Studio exposes a dedicated API endpoint, designed for direct consumption by AI agents and integration into existing inspection platforms, engineering systems, drone ecosystems, and enterprise workflows.

Built for the Agentic AI Era

AI Studio has been designed from the ground up as an Agent-first, API-based platform. It reflects eSmart Systems’ belief that enterprise software is shifting away from pre-built user interfaces toward dynamically assembled workflows powered by AI agents.

A Defensible Platform Built on Proprietary Intelligence

eSmart Systems’ competitive position rests on three pillars that no generic computer vision or foundation model vendor can replicate: more than a decade of proprietary, contextualized utility imagery and asset data built across 75+ utilities worldwide; patent-pending Adaptive AI that eliminates the customization barrier without retraining; and a platform architecture designed to absorb and leverage improvements in frontier AI models over time.

The result is an AI platform that enables immediate value and continuous improvement through rapid iterations, with performance previously unseen in these types of computer vision applications. An AI platform that is ready for the Agentic AI reality we are now in.

About eSmart Systems

eSmart Systems is the market leader in computer vision AI for electric utility infrastructure. Founded in Norway and operating globally, eSmart Systems serves 75+ utilities across transmission, distribution, and substation inspection — helping them reduce operational risk, improve reliability, and accelerate the transition to AI-powered grid management. Its flagship product, Grid Vision, industrializes AI-driven inspection at enterprise scale. AI Studio is eSmart Systems’ next-generation platform, opening its proprietary AI capabilities as modular, API-accessible services for utilities, technology companies, and inspection professionals worldwide.

For more information, visit ai.esmartsystems.com or esmartsystems.com

November 2025 – eSmart Systems has launched Verify AI, the first feature powered by its new and patent pending Adaptive AI technology. This advancement accelerates how utilities transform asset data into trusted insights that support safe and reliable grid operations.

Traditional AI has already improved digital inspection workflows. However, utilities still face high volumes of false positives, long deployment timelines, and difficulty tailoring results to local power grid requirements. These challenges impact decision-making and risk management.

Adaptive AI introduces a step change by enabling rapid value from limited or uneven training data. It adapts to each utility’s power grid and standards with minimal effort and improves continuously based on expert feedback.

Verify AI is the first feature based on Adaptive AI and is now available in the Grid Vision^® platform. Verify AI strengthens inspection workflows by reducing false positives and elevating only credible issues for review. This creates faster, more trusted results and allows inspectors and asset managers to focus on what matters most.

Adaptive AI will support additional features over time, including new defect detection and inventory capabilities, all designed to scale utility intelligence and improve operational outcomes.

Key benefits include:

• Faster deployment and value realisation
• Reduction in false positives and manual review effort
• Trusted insights aligned to each utility’s standards
• Stronger risk mitigation and prioritisation of critical work

Verify AI is available today in Grid Vision^®. More features powered by Adaptive AI will be released in upcoming product iterations.

About eSmart Systems

eSmart Systems is a global leader in AI-powered solutions for the inspection, analysis, and digitalization of critical grid infrastructure. Through our Grid Vision^® platform, we help utilities transition from manual inspections to intelligent, image-based asset management, creating a digital inventory of their transmission and distribution networks. This enables improved asset data quality, safer operations, reduced inspection costs, and extended asset life. With more than 20 years of international experience, eSmart Systems supports utilities worldwide in building smarter, more resilient grids for the future.

For more information, visit ai.esmartsystems.com and check out our announcement on the launch of AI Studio by eSmart Systems, incorporating our Adaptive AI technology.

As wildfire risk continues to escalate, utilities and regulators are under increasing pressure to reduce risk while maintaining reliability and affordability. This whitepaper brings together insights from leaders across the wildfire ecosystem to explore how integrated, prevention-first technology is enabling smarter decision-making and measurable results.

It examines how connecting intelligence across weather, vegetation, assets, and operations helps organizations move beyond reactive approaches toward proactive, data-driven, and cost-effective wildfire mitigation strategies.

Featuring:

• Don McPhail, VP Market Development, eSmart Systems
• Erik Åsberg, Chief Technology Officer, eSmart Systems

In today’s energy landscape, utilities are under pressure to do more with less, modernize aging infrastructure, respond to extreme weather, and meet rising expectations for reliability and sustainability. At the center of this transformation is data. But data alone isn’t enough. It’s what you do with it that matters. Join us as we go behind the scenes of eSmart Systems’ Grid Vision^® platform to explore how our intelligence layer is helping utilities turn raw inspection data into real-time decisions with AI, digital twins, and a smarter approach to grid management.

Don McPhail:
I’m joined today by Erik Åsberg, our Chief Technology Officer here at eSmart Systems. Erik has been leading the development of our AI and data platform strategy for over a decade and is widely recognized as a thought leader in applying advanced technologies to utility operations.
Erik, we often talk about Grid Vision^® as a game-changer for utilities. But behind the scenes, there’s a powerful data platform making it all possible. Can you walk us through what that platform actually does?

Erik Åsberg:
Absolutely. At the heart of Grid Vision^® is our intelligence platform, a cloud-native system that ingests, organizes, and analyzes all inspection data. Whether it’s drone imagery, infrared scans, LiDAR, or even handwritten field notes, the platform links everything to the utility’s assets. The result is a living digital twin of the grid.
But it’s not just about storing data. The platform transforms raw inspection inputs into structured, actionable intelligence. It gives utilities a single source of truth for asset condition, enabling them to shift from reactive maintenance to proactive, risk-based decision-making.

Don:
So it’s not just a database, it’s a decision engine?

Erik:
Exactly. Utilities have no shortage of data. The challenge is making sense of it. What makes our intelligence platform unique is that it brings together three key dimensions:

• A graph-based model that reflects how assets are physically and electrically connected, so every pole, conductor, and substation is contextually linked.
• Time-series history, which tracks how asset conditions evolve over time, enabling trend analysis and predictive maintenance.
• Spatial awareness, integrating GIS and environmental data to understand where assets are located and what risks they face, like terrain, vegetation, or weather exposure.

Together, these dimensions create a rich, intelligent model of the grid. This structure allows for fast, intelligent queries, like identifying aging assets with recurring defects, and enables seamless integration with AI, so insights can be surfaced and acted on in real time.

Don:
Let’s talk about that AI. How does the platform support the AI capabilities in Grid Vision^®?

Erik:
It’s AI-ready by design. Because the data is structured and fully connected, it’s ideal for training and deploying machine learning models. For example, our AI can detect defects in images and automatically associate those findings with the correct pole or component.
We’ve also integrated natural language interfaces. You can literally ask the system, “Show me all critical defects within 5 miles of a substation,” and it will return results instantly. That’s the power of combining structured data with modern AI.

Don:
That’s incredibly powerful. What about integration with other utility systems, GIS, EAM, BI tools?

Erik:
We’ve made openness a core principle. The platform exposes standard APIs and supports a query language familiar to many data teams. That means utilities can pull data into their existing systems without vendor lock-in.
And because we support spatial, temporal, and graph-based data, it’s easy to add new sources, like weather feeds or new sensor types, without rearchitecting anything.

Don:
Utilities are dealing with massive volumes of data. How does the platform support scale?

Erik:
Great question! It’s built on modern cloud infrastructure and designed to handle the scale and complexity of utility operations. Whether you’re managing 10,000 assets or 10 million, the platform can ingest, organize, and analyze that data in real time. As an example, we’re working with many large utilities across North America and Europe such as Evergy, Xcel Energy and E.ON.
And it’s secure, as demonstrated via our ISO 27001 compliance, and support of technologies such as SSO and multifactor authentication. Utilities retain full ownership of their data, including hosting data in their local region, and we provide enterprise-grade access controls to ensure privacy and compliance.

Don:
One thing I hear often is the need for faster insight-to-action cycles. How does the platform help with that?

Erik:
That’s one of its biggest strengths. Traditional inspection workflows are slow and fragmented. With our platform, everything is connected and instantly quarriable. You can go from image capture to defect detection to maintenance planning in hours, not weeks.
And because we support predictive analytics, utilities can start to anticipate failures before they happen. That’s the future: a grid that’s not just monitored, but truly intelligent.

Don:
Final question, what excites you most about where our technology and intelligence platform are headed?

Erik:
It’s the convergence of AI, digital twins, and real-time data. We’re building more than a platform, we’re building a foundation for autonomous grid operations. As we integrate more AI capabilities and natural language tools, we’re making it easier for utilities to interact with their data and act on it. This is only increasing in importance as we seek to make the grid more efficient and adaptable to extreme weather, changing supply side, and growth in demand.
We’re not just helping utilities inspect assets, we’re helping them reimagine how the grid is managed, making them more efficient in having AI take on the mundane and tedious work, and prioritize their investments where it’s needed most.

Don:
Thanks, Erik. For utilities navigating aging infrastructure, climate risk, and rising expectations, this platform offers something rare: clarity, speed, and control. That’s what real transformation looks like.

Want to learn more about how Grid Vision^® and our intelligence platform can help your utility? Book a demonstration or reach out to our team to see how we’re helping utilities turn data into decisions.

Featuring:

• Josh Allison, SVP Commercial, eSmart Systems
• Don McPhail, VP Market Development, eSmart Systems

In a recent address to federal regulators, North American Electric Reliability Corp (NERC) President and CEO Jim Robb described today’s grid reliability challenge as a “five-alarm fire.” He warned that while the reliability of the U.S. grid remains high, the risks to that reliability are intensifying because of extreme weather, growing demand, and aging infrastructure.

Among those risks, few are as destructive or costly as wildfires caused by electrical equipment failures.
In this discussion, Don McPhail and Josh Allison reflect on NERC’s warning and explore how AI-powered tools like Grid Vision® are helping reduce grid-related wildfire risk through proactive, data-driven insights.

Don: Josh, the NERC President called the state of grid reliability a “five-alarm fire.” When you look at recent wildfire events and infrastructure-related ignitions, that metaphor feels literal. What are you seeing from utilities right now? 

Josh: You’re spot on, Don. Utilities are living that reality. Drought, heat, and aging infrastructure have created a perfect storm. Many U.S. states, especially in the West, are facing unprecedented wildfire seasons. What we’re hearing is that utilities want to move from reactive inspection programs to predictive maintenance because every undetected defect or vegetation strike is a potential ignition. 

Don: The article mentioned “small-scale events and near misses” increasing. How does that tie into wildfire risk? 

Josh: Those “near misses” are exactly what we’re helping utilities identify earlier. A cracked insulator, a corroded clamp, or a frayed conductor might not cause an outage today, but in dry, windy conditions it can become an ignition point tomorrow. Grid Vision’s AI scans through vast image data to find those early indicators before they escalate, allowing maintenance teams to act on real risk instead of probability. 

Don: The NERC discussion also raised the need for better “shock absorbers” in the system, meaning tools that help utilities manage uncertainty. What does that look like in the context of wildfire prevention? 

Josh: Wildfire risk changes constantly with temperature, vegetation growth, and asset condition. AI becomes that “shock absorber” by continuously analysing new imagery and risk data. With Grid Vision, utilities can see which areas are trending toward higher ignition risk and proactively allocate inspection and repair crews before the fire season peaks. 

Don: The article also raised concerns around cost and affordability, that improving resilience can be expensive. How do utilities balance that with the pressure to reduce wildfire risk? 

Josh: Prevention is always cheaper than recovery. Through AI automation, we’re helping utilities cut inspection costs by more than 50 percent while covering more assets in less time. That efficiency allows them to expand wildfire mitigation programs without increasing operational costs, which is a win for both regulators and customers. 

Don: So, responding to NERC’s “five-alarm fire” is not only about building more infrastructure. It is also about understanding and managing what already exists more effectively. 

Josh: Exactly. Reliability starts with visibility. The grid is changing faster than anyone can rebuild it, but AI gives utilities the ability to monitor, predict, and prevent. That is how we put out the fire before it starts. 

The NERC warning highlights one clear truth: reactive maintenance is no longer enough. 

September 2025 – In a world where energy infrastructure must be smarter, faster, and more resilient, eSmart Systems and Equinor are leading a pioneering researching collaboration: the Infrastructure for Faster, more Accurate Analytics (IFAA) project.

Born from eSmart Systems’ deep roots in artificial intelligence and formally approved by the Research Council of Norway, IFAA is more than a research initiative. It is a step into the future of infrastructure inspection and defect detection. The project has already delivered transformative innovations, including Adaptive AI, a patent-pending technology that drastically reduces retraining time for AI models and enables real-time tuning based on user feedback.

Recognizing the project’s potential to establish a new benchmark for how utilities harness AI, eSmart Systems identified an opportunity to push the limits of model training and automation in real-world conditions. By developing smarter, adaptive systems that can continuously learn from data and feedback, the project aims to help energy companies achieve faster insight generation, greater operational accuracy, and more proactive maintenance strategies. Equinor joined as a strategic partner, bringing decades of operational expertise and a shared vision for AI-driven transformation. Together, the teams are exploring the frontiers of computer vision and applying innovations such as AI that learns from fewer examples, creating step-change improvements in efficiency and accuracy.

At the heart of IFAA is a commitment to open innovation. While each partner retains their intellectual property, learnings are shared across sprints, creating a dynamic feedback loop that accelerates progress and deepens insight. Both companies noting advancements in defect detection and AI-driven inspection automation. The collaboration now looks ahead to further expanding its impact by exploring varied use cases within each organization to drive even greater operational value.

Together, eSmart Systems and Equinor are turning AI into practical solutions that deliver value today while aiming to drive a smarter, safer, and more resilient energy sector for the future.

About Equinor

Equinor is an international energy company committed to long-term value creation in a low-carbon future. The company is headquartered in Norway with around 25,000 employees and with offices in more than 20 countries.

About eSmart Systems

eSmart Systems is a global leader in AI-powered solutions for the inspection, analysis, and digitalization of critical grid infrastructure. Through our Grid Vision® platform, we help utilities transition from manual inspections to intelligent, image-based asset management, creating a digital inventory of their transmission and distribution networks. This enables improved asset data quality, safer operations, reduced inspection costs, and extended asset life. With more than 20 years of international experience, eSmart Systems supports utilities worldwide in building smarter, more resilient grids for the future.

August 2025 – Evergy, a leading U.S. utility serving 1.7 million customers across Kansas and Missouri, has renewed its commitment to digital asset management through an extended collaboration with eSmart Systems. Building on the success of a 3.5-year digitalization and inspection program powered by Grid Vision^®, Evergy will continue to expand its digital asset strategy to create a future-ready, data-driven grid.

The original program transformed over 10,100 miles of Evergy’s transmission network into a centralized, visual asset repository, creating a full digital inventory of every transmission asset down to the component level. Over 1 million high-resolution images were linked to asset records, while 70,000+ annotated defects were cataloged to prioritize maintenance and repairs. This complete and searchable asset inventory has enabled engineering, real estate, and operations teams to access reliable, up-to-date data, dramatically improving data accuracy and situational awareness.

With this renewal, Evergy will continue to assess transmission inspections while expanding the use of Grid Vision^® to its distribution grid and digitalizing new asset types, including the full digitalization of inspections and imagery for underground vaults. By consolidating these asset types into a unified platform, Evergy will gain a deeper understanding of asset health across its entire network, enabling more accurate risk-based analysis and investment decisions.

Through this renewal, Evergy and eSmart Systems will continue working together to advance virtual inspections, enhance safety by minimizing unnecessary fieldwork, and support the transition toward a modernized, fully digital grid.

The continued collaboration will enable Evergy to expand its digital asset inventory, streamline inspection workflows, and unlock deeper insights to support the growing demands of a modern utility grid.

About Evergy

Evergy (NASDAQ: EVRG) provides electricity to 1.7 million customers across Kansas and Missouri. Headquartered in Kansas City, Evergy is a vertically integrated utility committed to delivering safe, reliable, and affordable energy. The company is actively investing in grid modernization and sustainable infrastructure to support a resilient, future-ready energy system. For more information, visit https://www.evergy.com/

About eSmart Systems

eSmart Systems is a global leader in AI-powered solutions for the inspection, analysis, and digitalization of critical grid infrastructure. Through our Grid Vision® platform, we help utilities transition from manual inspections to intelligent, image-based asset management, creating a digital inventory of their transmission and distribution networks. This enables improved asset data quality, safer operations, reduced inspection costs, and extended asset life. With more than 20 years of international experience, eSmart Systems supports utilities worldwide in building smarter, more resilient grids for the future.

12 August 2025 – eSmart Systems’ Grid Vision^® has been named a winner in the Best SaaS Solution for Energy, Utilities or Telecoms at The 2025 SaaS Awards.

Operated by global cloud computing awards body The Cloud Awards, The SaaS Awards judges have completed their final round of assessment of the finalists, resulting in winners being selected in each category.

The SaaS Awards celebrates outstanding levels of innovation and excellence in the software-as-a-service industry, from business process improvement solutions to cutting-edge uses of AI, and both niche and generalized services. The program received entries from organizations of all sizes across the globe, including the USA and Canada, the UK and Europe, the Middle East, and APAC.

The program will return to welcome new submissions in 2026, continuing to recognize the latest advancements in software-as-a-service.

To view the full list of winners, visit The Cloud Awards website.

Entries are now open for the long-running Cloud Awards program, which recognizes organizations at the forefront of cloud computing. Submissions are open until October 24, with earlybird entries accepted until August 29. Entries to other Cloud Awards programs, including The FinTech Awards and The Security Awards, both concluding in 2026, are also open.

About The Cloud Awards

The Cloud Awards is an international program which has been recognizing and honoring industry leaders, innovators and organizational transformation in cloud computing since 2011. The Cloud Awards comprises five awards programs, each uniquely celebrating success across cloud computing, software-as-a-service (SaaS), cloud security, artificial intelligence (AI), and financial technologies (FinTech).

Winners are selected by a judging panel of international industry experts.

About The Cloud Awards Program

The Cloud Awards identifies and celebrates the most innovative organizations, technologies, individuals and teams in the world of cloud computing. The program spans 36 categories, including ‘Best Cloud Infrastructure’ and ‘Best Cloud Automation Solution’. 

About The SaaS Awards    

The SaaS Awards focuses on software-as-a-service, with categories segmented into overall SaaS excellence, by sector, business operational processes, or outstanding uses of certain technologies (such as AI).

About eSmart Systems

eSmart Systems is a global leader in AI-powered solutions for the inspection, analysis, and digitalization of critical grid infrastructure. Through our Grid Vision^® platform, we help utilities transition from manual inspections to intelligent, image-based asset management, creating a digital inventory of their transmission and distribution networks. This enables improved asset data quality, safer operations, reduced inspection costs, and extended asset life. With more than 20 years of international experience, eSmart Systems supports utilities worldwide in building smarter, more resilient grids for the future.

With global electricity demand projected to grow by 4% in both 2024 and 2025*— one of the fastest growth rates seen in two decades — utilities face the pressing challenge of ensuring their grids can meet this surge reliably and sustainably.

Rising energy consumption and the integration of renewable energy sources make it essential not only to add capacity but to do so smartly and sustainably. We have seen first-hand there is one crucial element to meet this effectively: accurate grid asset data.  

Without this, utilities are essentially navigating blind. By understanding the true state of the grid’s condition, utilities can not only handle current loads but also plan smartly for future expansions and repairs. 

At the heart of effective grid management is accurate asset data. 

For utilities, maintaining a complete and up-to-date picture of their infrastructure is essential. Without it, effective operations management, future planning, and smart investment decisions become nearly impossible.  The first step toward improving the accuracy of grid asset data can begin with activities like grid inspections or a data quality improvement activities like grid inventory programs.

For either of these activities, utilities need to ensure that the data collection is focused on image-based asset data, collected with high-resolution cameras that capture detailed views of the infrastructure. These images, when processed through advanced tools like Grid Vision^®, provide a visual inventory of assets, cataloging them with pinpoint accuracy and linking the asset to additional information like inspection data. 

What benefits can you achieve with image-based accurate asset data? 

Once utilities start collecting and leveraging accurate image-based data, they can unlock several critical benefits that support operational efficiency and long-term grid resilience: 

1. Efficient Maintenance: Precise image-based data enables condition-based maintenance. This allows utilities to identify assets that require attention before they fail. This proactive strategy reduces outages, extends asset life, and prevents costly emergency repairs​. 
2. Better Investment Decisions: Accurate, consistent and objective insights of grid infrastructure condition allows utilities to prioritize capital investments based on risk. For example, virtual inspections enabled by AI can highlight critical areas where repairs are needed, helping to allocate resources where they will have the most impact​. 
3. Enhanced Planning for Growth: Accurate image-based asset data supports improved  forecasting for future grid expansion or reinforcement. As demand rises, utilities must not only repair existing infrastructure but also build new assets.  
4. Improved Safety and Compliance: Knowing the exact status of grid components ensures compliance and reduces the risk of incidents. This compliance aspect is particularly important as aging infrastructure and increased renewable energy integration strain the grid​. 

Data as the Cornerstone of Grid Management 

Accurate, reliable asset data is more than a convenience—it’s essential to managing current operations effectively and preparing for sustainable growth.  

By embracing advanced digital tools, like Grid Vision, and leveraging the power of AI, utilities can transform their data collection and asset management practices, ensuring they stay ahead of the curve in a rapidly evolving energy landscape​. 

*Source: International Energy Agency (IEA) report, 2024 forecasts on electricity demand growth. 

Want to see a demo of Grid Vision? Fill this form here