UX Designer
Bringing Clarity to Complex Space Data
UX Research + Design | NASA-Funded Astrophysics Mission
The Challenge
Princeton University, in collaboration with NASA, was developing a first-of-its-kind data analysis tool to support a space mission investigating the boundary of the heliosphere—where solar wind meets interstellar space. The project required a system to help scientists analyze complex, high-dimensional space data, critical to understanding space weather and solar phenomena.
However, the astrophysicists lacked the tools to analyze or visualize their data. Many had resorted to writing their own code—resulting in clunky, inconsistent solutions that were time-consuming and difficult to use. Princeton engaged us to design an intuitive, user-friendly interface that would enable scientists to explore complex datasets and generate publication-worthy visuals.
We can't visually represent the data more than we do now—that's the weirdness."
Stakeholder
Project Details
Timeline: 4 years; I contributed for 3 months in 2024
Role: UX Researcher & Designer; I redesigned complex workflows and built high-fidelity prototypes
Team: 12 developers, rotating UX team (6 researchers/designers), 1 PM, 2 scientist stakeholders
Client: Princeton University, in collaboration with NASA
Tools: Axure, Zoom
Deliverables: High-fidelity prototypes, workflow redesigns
Note: I was rotated off the project before the usability testing and implementation phases. This case study focuses on my contributions to research, design strategy, and prototyping.
The Challenge
Give space researchers an elegant and intuitive way to manage and reduce the dimensionality of their data so they can better analyze and interpret phenomena in space.
My Role
As a UX researcher and designer embedded in a cross-functional team of developers and rotating UX contributors, I joined the project midway through its multi-year development cycle. With another UX team member, I led a design sprint focused on reducing data dimensionality and enabling intuitive scientific exploration, collaborating closely with two astrophysicist stakeholders to shape and validate a new workflow for exploratory analysis.
Key responsibilities included:
-
Conducting contextual inquiries and usability observations
-
Translating scientific analysis needs into actionable UX flows
-
Designing and iterating high-fidelity prototypes
-
Collaborating with developers and PMs to align technical feasibility
Approach
Understanding the Domain
Designing for Dimensionality
.png)
Iteration and Refinement
Usability Testing
Understanding the Domain
To design an effective tool, we first had to understand the scientific analysis goals and constraints. Through interviews and observational sessions, we worked with astrophysicists to define:
-
Key data dimensions: time, energy, flux, pitch angle, and look direction
-
Required capabilities: flexible slicing and averaging of datasets, multidimensional reduction
-
Visualization needs: support for dynamic views beyond traditional x- and y-axis charts
In this domain, "dimensions" refer to scientific variables such as time, energy levels, and the directional orientation of the spacecraft's instruments. These variables interact in complex, nonlinear ways, which makes traditional plotting insufficient. Designing for dimensionality meant supporting exploratory workflows that reveal meaningful relationships across these axes.
We analyzed legacy mission data and existing tools to identify usability gaps and anticipate emerging needs-including the requirement to support future emergent data dimensions post-launch.
Insights
-
The affinity map based on our interviews led to six key themes and four priority insights we wanted to address.
-
How Might We questions narrowed the focus of our solution.
Affinity Map
Key Insights:
One
Essential Dimensions:
Scientists needed to visualize data across critical dimensions including time, energy, flux, look direction of the space instrument, and pitch angle. Additional dimensions would emerge once the spacecraft launched.
Two
Analytical Control:
They needed the flexibility to reduce the dataset by averaging over specific variables or slicing individual dimensions to explore patterns.
Three
Visual Flexibility:
Traditional 2D plots were not enough. Users needed to manipulate the data space more fluidly--beyond x/y plots.
Designing for Dimensionality
To support flexible data manipulation, we designed a streamlined flow that allowed scientists to:
-
Select dimensions for plotting (x-/y-axis)
-
Choose variables to average across
-
View individual "slices" of a dimension using a slider or table interface
This was delivered through a redesigned "Select Data to Plot" screen and an dynamic modal for adjusting dimensions mid-analysis.
Why this mattered:
These changes allowed scientists to interact with data in a way that mirrored their mental model of space-based phenomena—reducing friction and cognitive load.
Goal Statement
Our EdTech learning platform will let students practice academic content while encouraging interpersonal communication which will improve academic performance and interpersonal skills and engage students in the school environment.
Filling the gap
There were so many stellar academic products on the market already, so we realized that the best way to solve our problem was to create a layer of interpersonal activities that integrates with the online academic practice tools teachers already use to help students communicate and have fun with each other.
Select Data to Plot Screen
Prototype of modal with dimension plotting feature. Users select which variables to visualize or average using checkboxes prior to sending their data plots to the workspace.
First Version of "Adjust Dimensions" Modal
Once data plots are in the workspace, dropdowns emphasize data processing method over variable selection.
Interactive Workspace Prototype
With this model, users can use a slider to explore data slices across selected dimensions.
Iteration 1:
-
Emphasized ata display method first
-
Displayed all dimensions in dropdowns
-
Enabled limited exploration along one axis via slider
Iteration and Refinement
In collaborative working sessions, we discovered the original design needed greater granularity and navigational scalability. Based on stakeholder feedback, we:
-
Shifted focuse to starting with dimensions, then defining how each is handled
-
Introduced a table view to support selecting from 80 "Look Directions."
-
Replaced the static slider with a grouped slider (by 10s) for scalable navigation.
-
Future-proofed the UI for additional dimensions, ensuring it would scale as new data types emerge.
Dimensions-First Dropdown Interface
Users could select how each dimension should be handled (e.g. plot, average, or slice). A dynamic table then appeared for fine-tuning.
Combined Table + Slider
A final version merged the slider and table to accommodate high-volume datasets while maintaining clarity.
I haven't seen anything like this!"
Client Stakeholder
Wireframes
I created this rough wireframe and a lo-fi prototype to explore our concept through the teacher user flow.
For our initial mockups, we chose a calming blue theme for our design. These mockups follow the lo-fi wireframes closely.
Initial mockups
_edited.jpg)
Final Homepage Mockup
To simplify the usability, I condensed all of the main actions to the header.
Results
-
Improved clarity: Scientists reported a more intuitive experience compared to their self-built tools
-
Expanded flexibility: Enabled workflows not previously possible in their own tools
-
Validated concept: Design finalized and queued for implementation ahead of live mission data
Although I transitioned off the project before these phases began, I paired closely with a teammate who carried the design forward, while I remained available for consultation as necessary.
Key Feature: Reducing Dimensionality
Dimensionality Visualization: Joanie effortlessly manages dimensions like time, energy, flux, look direction, pitch angle, and mass.
Adjusting Dimensions: Joanie can average dimensions, select individual slices of dimensions to analyze, or average subsets of data, in addition to choosing which dimensions to put on the x-, y-, and z-axes.
Intuitive Interface: Our design supports both novice and expert users, minimizing the need for coding.
Outcome
The feature we developed for reducing dimensionality will greatly benefit the IMAP mission. Scientists like Joanie will be able to conduct advanced data analysis more efficiently and accurately. Our iterative, user-driven approach ensures the tool will meet all requirements and be ready for the 2025 launch.
Impact and Lessons Learned
This project required deep collaboration with domain experts in a highly specialized field. I grew more confident navigating complexity and working as a translator between science and UX.
Key takeaways:
-
Alignment is critical: When working in specialized fields, tight collaboration with domain experts is essential
-
Don't oversimplify the interface—simplify the decision-making
-
Tailored usability testing is non-negotiable: Validating a workflow in a specialized context requires crafting scenarios that reflect the actual scientific process
-
And, we would all absolutely go back to the beginning and rethink this design. Knowing what we know now, we would have structured and designed it differently from the start.
UI Add-On: Improving Feature Visibility
As part of internal design support for this project, I led a UI-focused redesign to address a critical usability issue. Without fail, scientists overlooked key tools in a left-hand toolbar. Using UX best practices and heuristic evaluation, I repositioned these tools to more intuitive areas of the interface, improving alignment with standard UI patterns and natural user workflows. Although I was not directly involved with the user testing, the redesign was tested with end users and implemented in the final design.