ExxonMobil Internship

Combine and redesign two applications to improve production engineers' well tests analysis experience.

Role

Date

Project type

Skills

UX Design Intern

May - July 2019

Personal Project

Interviews, Wireframe, Usability Testing, Task Flow, Prototype

Why

Commingled Well Test Analysis (CWTA) and Well Test Validation (WTV) were tools targeted for improving the accuracy of well tests, and have been identified as valuable tools to the business. CWTA and WTV had similar workflows and UI components which makes combining the two applications a possible solution to optimize users workflow.

Challenges & Goals

There were some challenges combining CWTA and WTV:

Restructured the workflow. Users felt confused about the current workflow, and they had to cross-reference 
other platforms to get the data they needed. My goal was to
make the new workflow intuitive and to reduce friction.

Presented information more effectively. Some of the charts were difficult to understand and users were not getting insights from them. My goal was to present information in ways users could easily understand, so they could get insights and make decisions quickly.

Process

Project Kickoff 

The project started with several kickoff meetings. During the meetings I had with stakeholders and scientists, I gained knowledge about the oil and gas industry, as well as gathered information about CWTA and WTV.

Previous Research 

The UX team has conducted usability tests on both CWTA and WTV before. Journey maps with users workflow, users feedback and team’s findings have been created. Read documentation from previous research helped me understand the known issues.

Content Audit

I conducted a content audit of the tools.  I listed the pages, data, graphics and plots in CWTA and WTV, and illustrated similarities and differences between these tools. This step allowed me to discover overlapping content and generated ideas for the combined workflow.

User Interviews

Objective

The purpose of the user interview was to help me understand production engineers' experience performing CWTA and WTV. After two remote interviews, I was able to understand their needs and pain points.

Main Questions

Can you tell me what production engineer does? What did you work on today?

Can you tell me when's the last time you perform CWTA and WTV?

What's the biggest challenge using CWTA and WTV?

How do you interact with the plots and trees?

Key Takeaways & Design Implications

Based on the feedback I got from the user interview, I determined some must-have features for the new design.

Takeaway

Design Implication

Users needed to go back and forth across different tabs for information. They also needed to open other platforms to check different data.

Put relative plots together and provided information which helps them make decisions.

Takeaway

Design Implication

Users were having many unexpected issues and their work got interrupted sometimes.

Allowed users to save their current work and they could come back to finish it later.

Takeaway

Design Implication

Users had a different preference for viewing the data. Some user preferred tables and numbers, and some of them believed plots bring them more insights.

Provided tables and plots at the same time. Users could switch between plot view and table view based on personal preference.

Takeaway

Design Implication

Users valued their time but they needed to spend more time to figure out how CWTA works.

Provided a visual step-by-step overview and guide users through the process.

Paper Prototype

I started with low-fidelity prototype so I can iterate rapidly.

Usability Testing

I scanned the paper prototype and put it into PowerPoint slides. By sharing control of the computer to the participants, I completed three usability tests. The primary objective of the test was to determine if my understanding of the new workflow was accurate through testing scenarios where users will decide to approve or reject two or more well tests.

Findings

After the usability tests, I created an affinity diagram of the data and derived frequency and severity of the different elements.

SME Reviews & Design Critiques

I showed my paper prototype to scientists to ensure the workflow is feasible. During our team's design critiques, I also got great feedback to improve my paper prototype.

Design Iteration 1

Issue

Participants had no idea what Case Manager is.

Chosen Solution

Created two tabs on the left, one for adding new cases,  another for Case Manager.

Pros

 - More flexibility. Users could 
   switch between tabs.
 - After the user clicked the Case
   Manager tab, they could 
   understand the function  
   quickly.

Considered Solution

Kept the welcome window as a pop-up window. In the window, allowed users to open recent cases and went to Case Manager.

Pros

 - Easy to open recent cases.
 - Gave the user more hints 
   about Case Manager.

Con

 - It might not be easy to 
   recognize cases just by case
   names.

Issue

Chosen Solution

The relationship between the overview plots and detail plots were not clear. Participants didn't understand the function of the time control bars and they didn't notice the legends.

Put the overview plots on the left and detail plots on the right.

Pros

 - Time scales were clear.

Con

 - The detail plots became
    smaller.

Considered Solution

Kept placing detail plots on the top, and overview plots down. Added a timeline with control bars on the top.

Con

 - The time scales of different 
    plots would be confusing.

Issue

Chosen Solution

Participants would like to know more details about each segment. In some cases, they needed to compare segments.

Used a drop-down list to select segments. Added a button for adding and comparing segments.

Con

 - When there were several 
   segments on the plot users 
   might not know the segment 
   number.

Considered Solution

Used the time control bars to view details.

Con

 - When there were several 
   segments on the plot, it's 
   tedious to adjust the time 
   bar.

Hi-fi Prototype

I used Axure to make the high-fidelity prototype with interaction.

Task Flow

Usability Testing, Stakeholder Reviews, and Design Critiques

In order to make the prototype more intuitive and easy to use, I conducted two remote usability tests.  The goal of the tests was to identify as many aspects of the prototype that could be improved. Discussion with team members and the stakeholder also helped me iterate the prototype rapidly.

Design Iteration 2

Takeaway

Users didn't always get a notification to validate well tests, they needed to go to another platform to see what new well tests were completed. 

The test duration was usually less than 24 hours so the time range needs to be more specific.

Change

Listed recent completed well tests so the users could open a case directly.

Made the time range more specific, includes hours and minutes.

Kept the previous design for advanced search.

Takeaway

More information could be provided here to help the user make judgments.

Change

Provided diagnostic for each segment, based on the diagnostic result, users could determine to edit or delete a segment.

Takeaway

Participants didn't notice the radio buttons, they didn't know they need to select a segment before they run the analysis.

Participants didn't notice the button.

Change

Added descriptions to inform the users what action they need to take.

Changed the radio buttons design into a drop-down list.

Users saw the segment number and test duration when they placed their mouse over a segment.

Takeaway

Participants didn't get many insights from the analysis report, they tend to skip this step.

Change

Simplified the workflow by showing two kinds of reports on the same page.

Visualized some important data from the reports. 

Final Design

Task Flow

Design

EC Record Overview

A table showed an overview of yearly well test results. By using icons presenting different test results, it saved users time from reading texts.

Past Cases

Users could open unfinished cases, opened old cases, and pulled out test reports from here.

Context Data & Diagnostics

Put the overview data and details on the same page, saved users' time from clicking different tabs. The system showed five kinds of plot as default. Users could select more plots from the drop-down list.

Compare test report

Users could pull out a similar case to compare with the current case. It helped users determine to send which case's data to the database.

Trending Analysis - plot view and table view

The plots visualized the data in the table. Users could review the data by their preference. I designed the bar graphs showing the difference between forecasted data and the test data. I also redesigned the legends to help users understand the plots easily.

linkedin.png
email.png

© Sherry Wu | 2020