Are we getting accustomed to the smart equipment and apparatuses that make up the smart grid? Are we indifferent about amazing systems that monitor, control and manage the power delivery system? It’s easy to be either way. Digital technology has been with us for decades. Each generation of these smart grid systems and schemes have brought improvements and increased capabilities. Smart grid technology has been modernizing the power grid’s infrastructure with a virtual infrastructure.
This virtual infrastructure is powered by equipment with sensors, transducers, monitors and lots of other digital electronics. They generate big-data that has brought about asset management systems, load management systems, demand management systems, etc., which has led to an infrastructure with awareness. Smart grid technology has matured and advanced. With each advancement, the technology has been revised and upgraded, and new applications have been able to leapfrog established technologies.
The result has been advancements that are astounding, and now that artificial intelligence (AI) has arrived the whole process has quickened, but there is a downside. How do we justify replacing the existing technologies that are no longer suited for their role in an industry that’s risk-averse and highly regulated? There are many smart grid applications that fall into this category, but let’s look at one in particular, smart meter technology.
A Valuable Tool for the Smart Grid
It really started becoming important to the power delivery industry as we prepared for the twenty-first century. Many of us began learning about advanced metering infrastructure (AMI) technology shortly after the millennium happening. Initially, it seemed like just another electronic meter, but the integration of ICT (information and communication technology) made AMI 1.0 unique.
It allowed two-way communication between smart meters and a central system. It also improved the quality of power consumption data. About the same time AMI had our attention, smart electronic devices were being adopted by our customers. IoT (Internet of Things) technology was also on the scene, and the behind-the-meter (BTM) world was getting wired.
This connectivity started changing things beyond the edge of the grid by giving customers remote control and monitoring abilities. Residential, businesses, and industries took advantage of the applications. AMI 1.0 gave the utilities access to the BTM environment, which changed both sides of the meter for the better.
About twenty years later AMI 1.0 was showing its age, its internal battery was at the end of its life and its error rates were climbing. Other issues suggested it was time to upgrade the application. Vendors quickly pointed out that AMI’s next generation was available for deployment, AMI 2.0 had new features and more capabilities that made the smart meter smarter.
It was time to trade up and start installing AMI 2.0. In a world where big-data was abundant, AMI 2.0 would better fit, with superior accuracy, It was also able to utilize what the customer had implemented. AMI 2.0’s networking offered utilities much needed advancements like high resolution data, peer to peer communications, edge computing, embedded computing power, etc. The industry listened and AMI metering became indispensable for modernizing the power grid.
By the end of 2023 (last figures available), smart electric meter penetrations had reached a 43% of the global electric meter market. That was somewhat over 1.06 billion units installed according to reports. Current projections indicate this technology has continued growing with higher penetrations and large numbers of installations, but there is a small hitch. Those 1.06 billion smart meters represent a jumble of technological metering arrays.
Some are pre-AM (i.e., electronic meters), others are AMI 1.0, and the rest are AMI 2.0. Adding to the challenge, vendors offered a wide selection of models ranging from economy budget deals to top of the line versions. That means many of those in-service smart meter systems offer limited value to utilities and grid operators, but they’re still using them. Many utilities are aware of the situation and are planning replacements, but that has started discussions about which ones will be replaced. There’s also the question, is there another choice?
Foggy Crystal Ball
Obviously the AMI 1.0 vintage are the chief candidates for replacements, but also under discussion is what to replace them with and when to do it. The AMI 2.0 systems of today have improved their performance and operability over those that were initially deployed. Various manufactures are discussing the next generation of AMI. A few have implied their systems incorporate features consistent with the next-gen’s, AMI 3.0, functionality.
AMI 2.0 evolution has been progressing since its initial implementation. The latest AMI 2.0 has improved its metering accuracy, its edge computing capabilities within the meters have been upgraded. The real-time interaction between customers and the utility has been advancing. In addition, AMI 2.0’s ability to push out updates for its software and firmware gives it an advantage. Could it be that AMI 3.0’s much anticipated appearance is already happening?
It certainly appears that the shift toward AMI 3.0 is underway. It’s not hard to find vendors advertising that their latest AMI platforms are enhanced by AI, which implies the AI feature is a minor upgrade. Nor is it unusual to find news releases saying their AMI systems incorporate AI as an integral part of its design, which fits the next level or AI-powered applications. That’s the heart of defining what AMI 3.0 is and how it works. It’s only natural for suppliers to say their AMI systems have next-gen features, and these statements are all valid. This is the usual progression as one technology evolves into the next.
Bear in mind, however, there are degrees of AI immersion into the next-gen AMI, which leads to discussions about potential improvements and additional features for AMI. Understanding AI will be crucial in shaping the future of AMI technology, just as it has been for many other fields. The combination of AI and AMI is improving AMI’s operational efficiency by being able to handle more data and datasets than previous AMI versions. It can also take advantage of AI’s ability to identify patterns, helping utilities recognize network issues, problems, and difficulties when the data is coming from an assortment of sources.
AMI with AI Upgrades
The simplest level of AI upgrades are the AI-enhanced applications, which primarily expand the overall system. The meters’ ability to communicate with each other can create innovative ways with a little help from AI improving the management of distribution networks in addition to improved operations and maintenance. Using AI-enhanced edge computing can send pre-processed data to the centralized cloud-based data analytics system instead of raw data, saving critical time.
Another AI-enhanced AMI 3.0 application assists the peer-to-peer functions to be more efficient. One example is automatic outage detection that can recognize distribution network problems in real-time. Many AMI meters have GPS (global positioning system) abilities and AI-powered AMI 3.0 uses it to establish exact location of problems, isolate them, and dispatch a crew without any human intervention.
The second level of AI upgrades is the AI-powered AMI 3.0 application. It uses AI as a key component of the partnership. Imagine a utility having thousands of meters transmitting big-data to utilities daily. AI-powered AMI 3.0 integrates real-time databases from weather to energy consumption along with others into a single source of information that’s capable of recognizing patterns and making predictions.
The third level of AI upgrades is the AI-driven AMI 3.0 application, which uses AI as the predominant force behind the AMI’s functionality. AI-driven AMI 3.0 is the perfect application for addressing the growing issues associated with bidirectional power flows produced from an assortment of distributed energy resources BTM. This application combines big-data from multiple databases and datasets to better control these bidirectional power flows, balancing sources with loads more efficiently.
An exciting AI-driven AMI 3.0 application being discussed can identify where vegetation is hitting power lines. Without going into a lot of detail, it uses the meter’s peer-to-peer function as well as GPS. By analyzing small variations in voltage and current waveforms it’s possible to determine that arcing is taking place when the vegetation hits the power line and GPS gives its location. This could be extremely valuable in preventing conditions that lead to wildfires.
The potential for AI upgrades to AMI are limitless and it’s a pretty good bet AMI 3.0 isn’t too far in the future. Looking at the examples above, there is a good argument that AMI 3.0 it’s already here, but there’s so much taking place that it hasn’t been labeled yet. The more we look into AI the harder it is not to see what a positive force it is when it comes to modernizing the power delivery system. Technological advancements are going to be uncomfortable and disruptive, but after using them, would you really want to go back? It’s the future, heck Wendy’s is using AI in their drive thru and both sides of the speaker are happier!
