The mission of this group is to bring together utility professionals in the power industry who are in the thick of the digital utility transformation. 


You need to be a member of Energy Central to access some features and content. Please or register to continue.


Artificial Intelligence Becomes Key in Utility Applications

image credit: ID 165707898 © Setiawanarief111 |

Artificial Intelligence Becomes Key in Utility Applications

Utilities find themselves collecting more information nowadays than ever before. Acting upon that information is challenging, but artificial intelligence is an emerging software area that could help with the process. Most utilities have limited experience working with the technology, so best practices are needed. 

Know Thy Data

Utilities do not want to simply collect information. They need to correlate it, see connections, and make changes to improve the business. 

So when working with information, it become important for them to understand the data and what it means to the business. That understanding often varies by department. For instance, the field service team would need customer information for location purposes, and the customer care team needs to monitor energy usage. When data is loaded into an AI application, such differences need to be accounted for. 

Start Simple

AI has vast potential but also prominent pitfalls. In collecting terabyte or petabytes of information, utilities have the potential of getting lost in the data, digging too deeply, and making the process of arriving at a conclusion challenging, sometimes impossible. In some cases, they enter too many variables or have vague potential outcomes. In addition, implementing, training, and evaluating a simple model takes much less time than a sophisticated one. Rather than boil the ocean, they should start with a simple tangible and reasonable metric, say reduced service time in the call center and build up from there.

Limit the Time Needed for Fine Tuning

The results from any AI model will never reach 100% because the tools are based on probabilities and leave room for something unexpected. Results in the 70% and higher range are often considered actionable. Sometimes, businesses try to raise the probability by tuning their AI model. Such quests chew up time, money, and resources. These projects, like all others, are usually time-constrained, so a wise approach to fine tuning seems to be needed.  Close enough to the original goal results often should be sufficient enough to label the project a success and move from examining your operation to improving it.

AI is an emerging technology that is changing how utilities build enterprise applications. The software has tremendous potential, but implementing it wisely requires that energy companies dig into their data at the start of the project, set reasonable objectives, and focus on delivering highly likely correlations rather than perfect ones.

Paul Korzeniowski's picture

Thank Paul for the Post!

Energy Central contributors share their experience and insights for the benefit of other Members (like you). Please show them your appreciation by leaving a comment, 'liking' this post, or following this Member.


No discussions yet. Start a discussion below.

Get Published - Build a Following

The Energy Central Power Industry Network is based on one core idea - power industry professionals helping each other and advancing the industry by sharing and learning from each other.

If you have an experience or insight to share or have learned something from a conference or seminar, your peers and colleagues on Energy Central want to hear about it. It's also easy to share a link to an article you've liked or an industry resource that you think would be helpful.

                 Learn more about posting on Energy Central »