Arizona Public Service: PowerBI Dashboards
Client
Arizona Public Service
Role
UX Researcher
UX Designer
Timeframe
9 weeks
Tools
Figma
PowerBI
Project Overview
An energy & utilities company, Arizona Public Service, needed to analyze their Revenue Protection Team’s current internal tools and processes for identifying energy theft to determine areas for improvement within their workflow.
Business Goal
Arizona Public Service needed to increase the number of energy theft investigations per day. Following a steady increase of customers, it has become challenging for a Revenue Protection team of 3 to keep up with the increase of investigations using their current process and tools.
Research
Value Stream Maps
We began working with the Revenue Protection team by creating a current state Value Stream Map to better understand the team’s process for identifying cases of possible energy theft. With the current state map we were able to visualize, analyze, and identify areas to improve the steps within their process. The future state Value Stream Map included our recommendations to maximize their available resources to ensure that materials and time are used efficiently.
Current State
Future State
Our future state recommendations (in blue text) will improve the Revenue Protection Team’s workflow by automating certain areas within their process. We concluded that creating a Scoring & Tracking database that automatically flags meters with events linked to possible energy theft will significantly reduce their time spent on investigations allowing the team to investigate more cases per day.
Using PowerBI to create the Scoring & Tracking database would be the most efficient option to deliver a platform to meet their data needs. However, the Revenue Protection Team currently has PowerBI dashboards that capture data similar to their needs. My next step is to figure out why the team is not utilizing their current PowerBI dashboards.
User Interviews
I conducted 3 individual interview sessions to observe the current process the Revenue Protection team uses to detect and assess possible cases of energy theft. These sessions helped identify areas for improvement within the Revenue Protection team’s current workflow and PowerBI dashboards.
Interview Synthesis
From the user interviews, I concluded that the following reasons are why the Revenue Protection Team is not using their current PowerBI dashboards:
PowerBI data is incomplete
Not enough filters to view the data needed
Dashboards do not include all meter and consumption events needed to identify theft cases
Define
User Flow
After conducting user interviews to gain an understanding of how the Revenue Protection team identifies cases of energy theft, I mapped out their user flow to better visualize the steps within their investigation process, gain insight into how and when each system is used, and highlight areas for improvement.
Current User Flow
Based on the information from the user interviews and current user flow, I concluded that vital data needed to be condensed into one reliable source to reduce time spent on each investigation.
Ideate
Wireframe Designs
Using the information from the user interviews and current user flow, I was able to create wireframe designs of PowerBI dashboards.
Mid-Fidelity Designs
After figuring out the layout of each page, I added context to the data tables, data visualizations, and data filters.
Mid-Fidelity Design Feedback
The Revenue Protection team liked the overall concepts of the PowerBI dashboards. The team provided specific data points they need to view in order to successfully investigate cases of possible energy theft. I organized each data point into one of the 3 categories: data tables, data filters, and data visualizations.
Prioritize
Data Mapping
Before creating a prototype with designs that incorporated the Mid-Fidelity feedback, I met with the development team to discuss the new data points. To start visualizing what the next iteration could look like, I created a dashboard that combined all data filters and all data table information on one page. The developers and I identified the source for each data point and prioritized each data point in regard to what is technically feasible for the first version of the dashboards.
Prototype & Test
Prototype
After prioritizing and mapping the data points from the Mid-Fidelity design feedback I started my next iteration of the PowerBI dashboards. I used Figma to create a prototype of the PowerBI dashboards. Each dashboard displayed data pivotal to energy theft investigations focusing on Red Flags, Meter Events, and Meter Consumption.
Prototype Feedback
The Revenue Protection team was happy with the data and workflow I presented in the prototype. Their only feedback was for me to figure out a way to include information about the premise they were investigating at the top of the Drill Down and Consumption pages. Since there were only a few data filters that they use on these pages to manipulate the data, I only included the necessary filters at the top of the page. This gave me room to include a summary of the premise that they were investigating at the top of the page with only the necessary information. All filters can be accessed by opening the ‘Filters’ panel on the right-side of the page.
Design Solution
For the final design, I added premise summary details and only kept the most relevant data filter, the date range slider, at the top of the page.
Since I removed the data filters from the top of the page, excluding the date range slider, they were now accessible through the ‘Filters’ panel on the right-hand side.
Final Designs
Red Flags
The Red Flags page is the main page of the report. It displays all the premises that have been run through the PowerBI Scoring & Tracking predictive model with an assigned Probability of Theft percentage.
Summary data of the premises are displayed in a table along with their geographical location and the red flags that have been assigned to that premise from the model. Users can use the Drill Down feature in PowerBI to access the other three pages to see consumption data, events, and other information about the premise in detail.
Premise Drill Down
The Premise Drill Down page displays a timeline of the consumption, red flags, and events at a premise. To show the details for a specific premise, either drill down to the Premise Details page from the Red Flags page or use the Premise ID filter in the All Pages section of the Filter pane on the right-hand side.
The date slicer at the top-right of the page can be used to filter the dates displayed on the timeline to a specific range for easier investigation.
Historical Consumption
The Historical Consumption page shows a timeline of the energy consumption at a premise by year. This page is also accessed via drilling down from the Red Flags page or by using the All Pages filter. Users can click on a particular year in the legend of the graphic to highlight the data for a particular year.
The Account Events at the premise are also displayed as a timeline below the consumption chart.
Heat Map Consumption
Users can view the consumption data of a premise as a visual heat map. Like the Premise Drill Down and Historical Consumption pages, this page should be accessed either by Drilling Down from the Red Flags page or using the All Pages filters.
The page displays the total energy consumption for each month for a selected year along with the standard deviation of the total kW/H for the year. Below, the energy consumption for the entire month is shown, with green squares representing relatively low energy consumption and red squares representing high energy consumption. The heat map breaks down energy consumption by the day by the hour.
Users can select a particular premise, month, and year using the slicers at the top of the page. It is important to note that if a user navigates to the Heat map Consumption page by Drilling Down from the Red Flags page, they must remove the Drill Down Filter in the Filter Pane on the right-hand side of the report to select a new premise using the slicer.
Conclusion
The scoring and tracking database that we created in PowerBI reduced the maximum time for energy theft investigations from 48 hours down to 15 hours. This allows the Revenue Protection team to investigate more cases of energy theft per day and maximizes the use of resources across the company.