Big data analytics for Oil and Gas Industries.
/summary
Quicklook into
the project
What we doing
Desiging experience for monitoring, analyzing and predicting big data to increase the cashflow and production for a Oil and Gas industries.
Key responsibilities
Resolving UX/UI issues assigned to milestones.
Contributing ideas and solutions beyond existing issues.
Delivering visual language to equip the community with tools in order for them to accomplish
desired goals in efficient ways.
Introducing the material and minimal design culture to the users who are mostly
familiar with SCADA design language.
Constant and close collaboration with the product contributors.
Conducting workshops to ideate.
Impact
Improved UX and increased engagement and efficiency of product by 52%
based on task completion rate
Scaled the platform to manufacturing and ESG.
1000x A key feature in design helped to undermine overlooked oil accounts
/the product
A SCADA with
analytics
SotaOG is an analytics tool for the oil and gas industry that uses AI and ML to quickly analyze large amounts of data. It helps companies monitor, visualize, analyze, and predict production and cash flow, allowing for informed decisions. SotaOG maximizes efficiency, reduces costs, and increases profits in a data-driven industry.
Business goals to address
To provide a comprehensive web app that meets the needs of the oil and gas industry
Enabling users to monitor, analyze, visualize, and predict production and cash flow in real-time.
To create a tool that can be used by a wide range of professionals, from drilling engineers to financial
analysts and operations managers.
To establish a reputation as a reliable and trustworthy provider of advanced analytics and machine
learning solutions for the oil and gas industry.
/the process
From don’t know
to do know.
We initiated the process of outlining research methodologies that would effectively uncover the problems we face and their corresponding solutions. As a startup pioneering unique features in the market, our focus primarily rested on secondary research methods. Nonetheless, we organized brainstorming sessions involving the CEO and petroleum engineers to gain deep insights into the daily workflows of technicians, engineers, and CFOs within a facility.
Quick note: + has detailed study. Due to nature of this project cant include all study
Discovery
Explore the problem space, gather insights, and understand the context.
Project Vision Board
Stakeholder Interview
Technical Learning
Focus group & Skills mapping
Needs Evaluation
Heuristic Evaluation
Competitor Analysis
Insights
Learned about the business modal and how the structure works
Predicting cash flow is challenging due to market price volatility and unplanned production downtime.
Operators struggle with siloed systems for production tracking and maintenance schedules.
Competitors provide dashboards but lack robust predictive capabilities integrating financial and production data.
Define
Synthesize research findings, identify key problems, and create a clear direction.
Persona Development
Journey Mapping
Pain Points Identification
Stakeholder Review
Insights
Needs real-time insights into cash flow and production metrics
Requires alerts for equipment maintenance to avoid production halts.
Disconnected data sources causing delays in decision-making.
Difficult-to-use interfaces that require extensive training
Combine financial and production data for actionable insights.
Develop
Ideate, prototype, and validate the solution, suggest whats missing.
Features Conceptualization
Brainstorming Session
Technology Identification
Usability Review
Information Architecture
Stakeholder Review
Solution Design
Insights
A predictive model combining cash flow and production data adds significant value.
Interactive dashboards need to balance simplicity with access to detailed insights.
Mobile-friendly interfaces are crucial for production managers on the field.
Users appreciate customization options for reports and alert thresholds.
Unique challenges
Data Integration and Management
Product must be able to integrate data from multiple sources, such as production and
financial systems, to provide a complete view of production and cash flow.
Scalability
As the number of users and facilities grows, the design must be able to scale to
accommodate increasing amounts of data and more advanced analytics capabilities.
Visualization
Design must present large amounts of data in a clear and intuitive way, making it easy for
users to understand and make decisions based on the information. The challenge is t
find the right balance between simplicity and comprehensiveness, and to design
‘visualisations that effectively communicate key insights.
Identified pain points
Struggling to optimize production efficiency
There is a constant challenge in identifying and implementing improvements to maximize production output and minimize costs.
Difficulty in understanding product trajectory through historical data
Analyzing historical data to predict future product trends and performance is complex and time-consuming.
Tedious process of reconciling
tank tickets
The manual and error-prone process of matching tank tickets with run statements slows down reconciliation and increases the risk of inaccuracies.
Facing limitations with
SCADA systems
SCADA systems often lack the ability to visualize and analyze real-time data effectively, hindering timely decision-making and optimization.
/the design
Taking everything
to artboard
We need to design with two key points in mind: not all users are familiar with advanced dashboards, but most technicians are accustomed to SCADA tools, so a drastic shift to a modern interface may hinder adoption. Speed is critical; as a startup, quick market entry is essential. Leveraging IBM's Carbon Design System is a smart choice—it's enterprise-compatible, robust, and accessible.
pain point 1
Struggling to optimize
production efficiency
With vast amounts of data being generated, it's challenging to visualize and analyze it in real-time. This leads to delays in identifying optimization opportunities and collaborating effectively across teams.
I focused on highlighting critical data to help users prioritize actions. Providing data summaries reduces the time needed to make decisions. By pinpointing where to focus first, we minimize cognitive load, making interpretation and action more efficient.
pain point 2
Difficulty in understanding product trajectory through historical data
The abundance of sensor data generated every minute presents a significant challenge in understanding the historical trajectory of products. Manually sifting through this data to identify anomalies or trends is time-consuming and burdensome. It hampers the ability to optimize processes based on historical insights.
Although showing all data on table with alerts and insights is a quick solution its also wont work when looking over thousands of columns best way to show history in chunks of few months and highlight if any day needs attention make it as an alert .
I have introduced a history bar which can be scrolled to any date and it visually inform the user on which day there is ticket un matching, because we identified users having trouble identify un matched tickets with just the table sorting and filtering .
/what i have done differently
pain point 3
Tedious process of reconciling tank tickets
Reconciling tank tickets with run statements is a laborious task, especially when dealing with misplaced or misshapen tickets. These discrepancies can lead to errors in payment calculations and prolonged reconciliation processes. Finding alternative solutions to streamline this reconciliation process is essential to alleviate this pain point.
I designed a digital tank ticket reconciliation tool enabling users to upload or scan physical tickets using OCR (Optical Character Recognition). The tool automatically matches ticket details with run statements, flags discrepancies, and suggests corrections. It features a clear, user-friendly interface for manual adjustments, with visual highlights to expedite reconciliation. A searchable archive of past reconciliations further streamlines the process, saving time on misplaced tickets.
/failed attempt 1
This was my starting point: I organized all the data and designed a table-accordion to hide everything except mismatched entries. However, this approach led to long scrolling, with critical information spread across the table, making it time-consuming to add or delete tickets.
/failed attempt 2
I tried organizing the accordion under each well/facility name but ended up with too many clicks to add, remove, or verify against the reconciliation file.
When I failed each time it gives much better clarity to do what next.
/quick note
/failed attempt 3
I settled on a simple table, clearly highlighting critical data with all CTAs upfront for quick action. However, each cell takes up significant space, and when a facility uploads all tickets, the table becomes excessively large, requiring users to scroll extensively.
/the one that made sense
1000x
It’s a hit. Thousands of barrel of oil payment reconciled after the integration of Auto Ticket. Saving tens of thousands of dollars each month.
/impact
pain point 4
Facing limitations with
SCADA systems
Many modern industrial facilities use SCADA systems, but these have limitations, such as a lack of real-time 3D visualization, difficulty in simulating future scenarios, and limited integration with advanced analytics. A 3D facility design or digital twin model addresses these issues by providing an interactive, real-time representation of the facility, enabling predictive simulations, better decision-making, and seamless integration with IoT and AI-driven insights.
I designed an interactive 3D facility visualization tool integrated with SCADA data for real-time insights. It features live metrics, color-coded anomaly alerts, and drill-down options for detailed analysis. This solution overcomes SCADA limitations, enhancing clarity, streamlining anomaly detection, and enabling better collaboration and decision-making.
/version 1
I tried to clone SCADA, but it lacked real-world context, making it difficult to identify machines or valves from pictorial representations.
/version 2
This provides a more realistic view of the facility, closely resembling a blueprint, with equipment designed to look like real-world counterparts. It significantly improves usability compared to a 2D view.
/version 3
After evaluating the 2D view, I designed and developed a 3D facility view, greatly enhancing usability for understanding the facility layout and identifying faulty sensors. This helped engineers and technicians quickly pinpoint issues and find solutions.
/final_one_v4_hgdasdgs :)
Building on the 3D view, I developed a digital twin model of the facility integrated with real-time data. This reduced cognitive load for identifying production needs and resolving issues. It improved usability by optimizing oil and water flow without relying on third-party digital twin models. Engineers can now monitor machine performance from anywhere.
/the result
Design meets
key success metrics
The design ensures success by streamlining workflows, improving data visualization, and enhancing decision-making for users like technicians, engineers, and CFOs. Key metrics include reduced reconciliation time, faster anomaly detection, higher user satisfaction, and increased adoption of 3D facility visualization tools integrated with SCADA data, ultimately driving operational efficiency and strategic value.
Few clients reported increased profitability due to proactive decision-making, and the new History and Reconciliation designs received positive feedback from all clients.
Task Success Rate:Goal: ≥ 80% success rate. (e.g., generating a cash flow report, predicting
production trends).
Time on Task: Reduce time by 37.2% compared to last design. (e.g., matching sensor data with tickets)
Feature Usage Frequency: High usage of predictive models, report generation and reconciliation
Time Spent on Dashboard/Task: ~3 minutes compared to 8.3 minutes before
Error Recovery Rate: ≥ 95% recovery success.
Adoption Rate: ≥ 75% adoption rate. (e.g., History bar, reconciliation and 3d facility view)
Qualitative Feedback: User testimonials and open-ended feedback from usability tests.
/fun part
Learning Blender
and Three js
My favorite part of a project was replacing 2D images with 3D digital twin models using Three.js and Blender. I thoroughly enjoyed creating the 3D models and learned a great deal in the process. Since there were no Three.js developers available at the time, I took on that role as well.
It was impossible to overlook, and I knew it had to be implemented. With no other option, I took it upon myself to make it happen.
/learning
Minimal design is
not always user friendly
Data visualization for enterprise analytics emphasizes that overly simplified visuals can obscure critical details. In complex domains like oil and gas, users rely on rich, layered information to make decisions. Stripping down dashboards to achieve minimalism may hide essential metrics, trends, or alerts, forcing users to dig deeper for insights.
Effective design balances clarity with depth, presenting critical data prominently while allowing detailed exploration through intuitive interactions.
/take this case
Here I was, struggling to design an operational envelope (an operational envelope is a data plot used to identify high efficiency across multiple factors). These are a few iterations I made to plot the operational envelope. I made several mistakes, such as showcasing all the data at once or creating overly minimal plots that didn't convey enough information (like plots 2 and 3). After analyzing all the available data, I realized that certain variable data needed to be highlighted for a more in-depth understanding and effective decision-making.
/final plot
This plot stands out for its clarity and decision-focused design. It effectively uses color-coded markers, annotations, and a well-organized layout to present complex data in an easily interpretable way. Key decision points are highlighted, and critical constraints are visually distinct, guiding the user's focus. Interactive elements like tabs and drill-down options ensure progressive disclosure, reducing cognitive overload. Overall, the plot balances accessibility, context, and usability, making it ideal for informed decision-making.
The project was complex, and met almost all matrices, but there was room for improvement, Real time sensor input data and collaboration. Users gave positive feedback, but there were also negative feedback points to improve upon. Overall, the I found the project challenging and rewarding.
Thank You!
That’s it on analytics.
Do you like to check my next case study on
improving messaging experience?
Reinventing messaging
experience for Quail.
Read more
I would love that!
Lets go!!
Sounds encouraging
