Utilities are kind of important, there aren't many places in the world that don’t have some form of water distribution, electricity, or natural gas transmission. For the most part, these services have changed over the years to better accommodate those who are using them. These improvements have oftentimes made utilities easy to access, cleaner and safer. In almost every case these improvements have been highly beneficial for the governments and businesses implementing them and the customers using them. This leads many to look at where the next improvement will come from. In this case, it is artificial intelligence.
When the concept of AI using image recognition software comes up, it can conjure conspiratorial ideas of shady businesses tracking where people go, police states scooping people up off the streets, and other Orwellian ideas. However, in the case of utility companies, it’s anything but that. In the real world, AI-controlled drones learning from gathered images are aiding in the discovery and solution of potential utility problems.
In fact, in a white paper report conducted by Burns McDonnel, a construction engineering company, found that the concept of AI-controlled drones using image recognition software could have great benefits for utility companies.
Improving Recovery Efforts
The report done by Kyle Lundeen and Daniel Sierra found that when AI is applied to image recognition software it was able to make things easier for companies and employees alike. One of their primary examples was how AI could scan an area and evaluate the situation after natural disasters occur. It was explained that AI could learn where and what utilities are providing the most aid in an area, figure out how much time and resources it would take to get it back up and running, and how safe it would be to even approach the area.
They explained their point by elaborating that in 2017 Pureto Rican utilities had been hit by Hurricane Maria. That authorities did their best and that although they did nothing wrong the situation could have been greatly improved by AI. It was stated that “had utilities there used camera-enabled drones to conduct baseline aerial surveys of service areas prior to the hurricane, they could have later sent the drones back into the skies to survey and assess the damage. By comparing images before and after, AI could have identified everything from missing electric poles to chemical spills.”
Lundeen and Sierra’s work also said that AI controlled drones didn’t have to be regulated to cleanup details that they can serve other purposes like how, “AI could have then been enlisted to estimate needs for replacement poles and conductors, and to link to vendor databases to check availability, place orders and prioritize repairs.” What they mean by this is that AI can help humans repair utilities at a more efficient route, making it safer and easier. AI is saving time and money, while making sure employees are out of harm's way, which is almost always a good thing.
Another example from the report outlining the usefulness of AI to utility business was using AI controlled drones to make initial surveys of wildlife and vegetation. It was explained by Lundeen and Sierra that, “image recognition and AI tools can also help prevent outages by estimating vegetation growth rates and better predicting trimming needs.” It went on to say that, “in addition to maintenance requirements, these tools can be used to estimate labor hours and materials for these tasks.” This is another cost effective way that AI is aiding utility services across the industry.
What to take from this is that AI image recognition is making life easier for utility companies and those they employ. This technology isn’t just for electric companies but for all utility companies. By figuring out the best time, place and equipment to use in any given situation it means less trips for constant inspections, and less time for newer employees to get adjusted to their jobs. There is no real risk of unemployment, just a push to make everything more efficient.
AI isn’t just using drones to aid utility companies they’re also helping figure out the best distribution processes for an optimal power grid. It was explained in Digitalist Magazine that AI can revolutionize the utility industry, but some of the biggest aspects it can impact was demand management.
It was mentioned that the UK based energy company Upside Energy has been using machine learning based AI to manage their demand and storage needs in order to better support their grids. The AI they have been utilizing is called Open Energy described as an AI that, “controls devices with flexibility in their energy consumption to shift demand in real time.” What it does is learn where power is needed the most and puts it there. Instead of the same amount of energy going to all areas at once this system will prioritize certain areas at specific times based on what it has learned.
It is even estimated that machine learning could unlock, “up to 6GW of demand-side flexibility, which can be shifted during the evening peak hours without affecting end users.” This is allowing for better optimization, and less strain on the grid itself resulting in less blackout and overall a better quality of life for grid operators and customers.