Finding Real Partnerships: How Utilities Evaluate Artificial Intelligence Providers

The energy world is undergoing a massive transformation, rethinking systems designed more than a century ago to make way for the rise of smarter, cleaner technologies. It’s an exciting time — virtually every industry is becoming electrified in some way, electric vehicles (EVs) are gaining market traction, and there’s an active transition in support of Distributed Energy Resources (DERs), “small energy resources” typically found near Locations are located of electricity consumption, such as solar panels on the roof and battery storage. The latter is a big deal, and as the International Energy Association (IEA) points out, the rapid expansion of DERs will “transform not only how electricity is generated, but also how it is traded, delivered and consumed.”

To an observer, all of these changes are positive, lasting, and long overdue. But in practice, the rapid acceleration of renewable energy and electrification is adding stress and pushing the limits of our power grid. In addition to the pressures of renewable energy, the world’s energy systems are also facing critical challenges from extreme weather events related to ongoing climate change – droughts in Europe, heat waves in India, severe winter storms in the US – all leading to an exponential increase in inspections and maintenance , and repair costs. Utility leaders are now focused on improving grid modernization, reliability and resilience.

Take a picture, it will last longer

For utility companies, their equipment is often their most important asset and requires constant, careful maintenance. Performing this maintenance depends on a steady stream of data (usually in the form of images) that utility companies can analyze to detect operational anomalies. Collecting this data comes in many ways, from drones and fixed-wing aircraft to line workers physically walking the job site. And with new technologies like UAVs/drones and high-resolution helicopter cameras, the sheer amount of data has increased astronomically. We know from our conversations with many utilities that utilities are now collecting 5x to 10x the amount of data they have been collecting in recent years.

All this data makes the already slow inspection work cycle even slower. On average, utility companies spend the equivalent of 6-8 months of man-hours per year analyzing inspection data. (Provided from a customer interview with a West Coast utility company that collects 10 million images per year) A major reason for this surge is that this analysis is still largely done manually, and when a company collects millions of inspection images each year , the process becomes completely unscalable. In fact, analyzing for anomalies is so time-consuming that most data is already out of date by the time it’s actually checked, resulting in inaccurate information at best and repeated inspections or dangerous conditions at worst. This is a big problem with high risks. Analysts estimate that the energy sector loses $170 billion each year due to network outages, forced shutdowns and mass disasters.

Building the utility of the future with AI-powered infrastructure inspections

Making our grid more reliable and resilient takes two things – money and time. Fortunately, this is where new technology and innovation can help streamline the inspection process. The impact of artificial intelligence (AI) and machine learning (ML) on the utilities sector cannot be overstated. AI/ML is right at home in this data-rich environment, and as the volume of data grows, so does the AI’s ability to transform mountains of information into meaningful insights. According to Utility Dive, “there is already broad consensus in the industry about this [AI/ML] has the potential to identify failing devices much more quickly and securely than the current method, which relies on manual inspections.

While the promise of this technology is undeniable, building your own custom AI/ML program in-house is a slow, labor-intensive process fraught with complications and obstacles. These challenges have prompted many utilities to seek additional support from outside consultants and vendors.

3 things to consider when evaluating potential AI/ML partners

When looking for an AI/ML partner, actions speak louder than words. There are plenty of smart companies out there that could promise the moon, but utility executives should take a closer look at a few key metrics to accurately assess the impact. Key ones include how the vendor describes/delivers:

Growth of the model over time – Creating disparate datasets (data with many anomalies to analyze) takes a long time (often several years) and certain types of anomalies do not occur frequently enough to train a successful AI model. For example, it can be difficult to train an algorithm to detect things like rot, woodpecker holes, or rusted mufflers if they’re not common in your area. So don’t just ask the AI/ML provider about the quantity of their data sets, but also about their quality and variety.

Speed ​​– Time is money, and any reputable AI/ML vendor should be able to clearly show how their offering speeds up the inspection process. For example, Buzz Solutions has partnered with the New York Power Authority (NYPA) to provide an AI-based platform designed to significantly reduce the time required for inspections and analysis. The result was a program that could analyze asset images in hours or days instead of the months it had previously taken. This time saving allowed NYPA maintenance groups to prioritize repairs and reduce the risk of failure.

Quality/Accuracy – In the absence of real data for AI/ML programs, companies sometimes supplement synthetic data (i.e. data artificially created by computer algorithms) to fill gaps. It’s a popular practice, and analysts predict that as early as 2024, 60% of all data used in the development of AI will be synthetic (rather than real). But while synthetic data is good for theoretical scenarios, it doesn’t work well in real-world environments where you need real-world data (and human-in-the-loop intervention) to self-correct. Consider asking the vendor about their mix of real and synthetic data to ensure the split makes sense.

And remember, the work doesn’t end once you’ve chosen your partner. A new idea from Gartner is to hold regular “AI Bake-Off” events — described as “quick, informative sessions where you can see vendors side-by-side using scripted demos and a shared dataset in a controlled environment.” to evaluate the strengths and weaknesses of each. This process establishes clear metrics directly related to the scalability and reliability of the AI/ML algorithms, which are then aligned with the utility’s business goals.

Driving the future of the utilities industry

From more efficient workflow integrations to sophisticated AI anomaly detection, the utilities industry is on a far better path than it was just a few years ago. However, this innovation must continue, especially as mandates for T&D inspections are set to double by 2030 and the government has announced the maintenance and defense of energy infrastructure as top national security priorities.

There is still work to be done, but one day we will look back on this time as a turning point, a moment when industry leaders stepped in to invest in the future of our energy grid and lead utilities into the modern age.