How to Be Proactive with Your Data
A data-driven approach is what most utilities need now to avoid power outages. Vice Presidents of Engineering and CEOs of electric utilities want to know, where is the next outage coming from, how can it be avoided and how much money would the outages cost me?
They don’t want a power outage that leads to investigations. Several outages have recently resulted in more than just an inconvenience to customers. Power outages are now equating to double digit deaths in some areas and billion-dollar losses in the US. Therefore, utilities are quickly getting their data together to figure out the all-important question, “How do we get ahead of the next outage?”
Where Is the Data?
The data is available. Some would say there is too much data. However, how to get the right data out of which systems is always a debate. The departments all want to look at certain small specific situations that may not really help predict future outages
So How Much Data Do I Need?
The answer: It depends on who you ask.
A data platform company will tell you to load all of your unstructured data into their system. Most platforms cannot help guide you through which data will garner the best results.
Therefore, if you just set up a data lake, you may drown in your project faster than you anticipated.
Employees are constantly being asked to generate new reports or presentations with the data. Employee frustration can be high since they may be only one or two people reviewing the data. This makes it difficult to develop a data-driven culture if the other employees don’t understand the origin of the data points. Other employees may believe the data that is being used is old or outdated. So before you start your journey be specific about what you hope to achieve.
The More Data the Better, In Most Cases
Aerinet Solutions usually recommends about 5 years of data outage information to start. Why so much data? Data specifics matter when building your outage data points. The bigger the data set, the better the opportunity to find different issues, trending, commonalities, and to implement machine learning-Artificial Intelligence.
Most organizations are concerned that the data quality is not very good. If you’re not quite sure about your organization’s data quality, take a look at the article called Is It Time to Improve Your Data Quality?
What you may find is that you do in fact have a ton of data, but it is difficult to organize, standardize and qualify. The quick checklist will give you an idea about how you feel about your organization’s current data situation.
Where To Get Started
The first thing you need is the experts who understand the outage world. These experts can help define the best data fields for starting your project. If however you are guessing that your data quality may not be the best, don’t worry. Some of the smaller companies have Data Quality Management Programs (DQMP) that are available to help customers clean up their data and teach their employees about a data-driven approach.
What Is A DQMP and Why Do You Need One?
A DQMP helps organizations set up their data so that as new systems are added the coding and parameters are set up properly. This allows new data to be integrated easily. They also look at creating a standardized process for future use. More importantly, they teach the rest of the company what a data-driven approach is and how to use the system to forecast future outages.
Warning: Not every vendor you work with will view this as a recurring program but more of an opportunity to charge for professional services. Customers have not built out their data quality program because they don’t learn how they can benefit from the overall approach.
By having a program, all employees learn about the data collection, quality, security and processes to believe in the data. This holistic program creates confidence and a continuum for all employees to standardize the coding and classification of the data. A documented end-to-end process accompanied by an employee program, makes everyone a data driven employee.
You Need the Right Partner
At Aerinet Solutions, we recommend that the customer begin with 10 core data fields. These 10 fields are the foundation for future predictive outage maps. All of the data points are securely transmitted. By keeping processes secure, the data governance standards are met while still providing an on-demand data solution for customers.
It Won’t Break the Bank
Most utilities feel that they just don’t have the budget or are intimidated with tackling the data issue. Several of the small to midsize utilities have successfully loaded their data into the Aerinet-PwrMetrix platform without ever exceeding their budget. Using the PwrMetrix system, many utilities have been able to accurately predict future outages. The savings on these outages far outweigh the costs of the PwrMetrix. These tools cost significantly less than many of the larger companies’ enterprise systems, who try to convince the utility to switch out the entirety of the utilities existing systems. With PwrMetrix you can use your existing AMI, OMS, AMR and SCADA system data without having to pay additional connector fee costs.
Related Links below
What is Data Quality Management Program?