I helped design and lead research sessions to develop novel visualization programs for two ongoing projects at the NASA Jet Propulsion Laboratory (MOSAIC autonous networks and hypervelocity gas sampling in the upper atmosphere of Venus) and one at Caltech (Oman Core Drilling project).
I worked in a team of two designers and three computer scientists during a collaborative visualization program between JPL/Caltech/ArtCenter. We focused on levraging design thinking and visualization strategies to tackle current science challenges and help researchers discover new patterns in their data. I will highlight the research from the Oman Core Project.
How can we operationalize kilometers of short-wave near-infrared microspectroscopy core images to facilitate research on the creation and modification of the oceanic crust and mantle?
The Oman Core Drilling Project is a world-wide effort to understand the cration and modification of the oceanic crust. At the site in Oman, 3.2 kilometers of core samples have been recovered. Microspectroscapy researchers at Caltech needed a way to visualize thousands of core samples, identify patterns, share findings, view mineral mappings from the perspective of “non-experts” and streamline their tedious analysis and workflow procedures
To understand the challenges of the researchers, we designed multiple sessions to observe their workflow and identify other groups of scientists who would be using the visualization tool. We did a “know” and “don’t know” exercise after our first interview to determine how to proceed with different methods and questions.
01. Researchers have no way to organize and keep track of their analysis of each individual core sample, or identify patterns between samples.
02. The limitation of current software restricts the amount of colors that can be mapped to each mineral, making finding relationships within mineral maps challening.
03. Scientists who are not familiar with spectroscopy and mineral maps are skeptical of the data and unsure on how to interpret it.
04. There are no ways to navigate through the 4,000 samples and understand “at a glance” what areas are interesting to investigate.
05. Sections were marked as “interesting” through an analysis done by hand, which is prone to errors and bias.
After gaining an understanding of the problem space, we had multiple design sessions with researchers to test out quick ideas and sketch new concepts together. This informed further ideation and the formation of resources to use in paper-prototyping.
We develoed low-fidelity paper prototypes to get a quick feel for the high-level functionalities needed for the tool. We tested the design with our two identified user groups (expert spectrocopists and scientists with no spectroscopy knowledge).
To understand the challenges of the researchers, we designed multiple sessions to observe their workflow and identify other groups of scientists who would be using the visualization tool. We did a “know” and “don’t know” exercise after our first interview to determine how to proceed with different methods and questions.
01. Researchers have no way to organize and keep track of their analysis of each individual core sample, or identify patterns between samples.
02. The limitation of current software restricts the amount of colors that can be mapped to each mineral, making finding relationships within mineral maps challening.
03. Scientists who are not familiar with spectroscopy and mineral maps are skeptical of the data and unsure on how to interpret it.
04. There are no ways to navigate through the 4,000 samples and understand “at a glance” what areas are interesting to investigate.
05. Sections were marked as “interesting” through an analysis done by hand, which is prone to errors and bias.
From our previous interview insights and information from the paper prototyping sessions we narrowed down on current, core, desired, and beyond scope functionalities for the design. I drafted documents detailing these functionalities in depth, and presented them to a group of stakeholders consisting of the core Caltech research team, NASA JPL HCI supervisors and mentors, the engineering team, and external reviewers. This set of functionalities was central to understanding the main features to prioritize in development, and which features the team could push to a future development cycle.
1. Display 3 minerals at once
2. Stitch together core sections (by hand or digitally)
3. Individually analyze and store notes
1. Display multiple minerals
2. Global and local spatial views
3. Analyze spectra
4. Filter and search based on target (mineral type, depth/section)
5. Mark uncertainties and discrepancies
1. Consolidation of core logs
2. Trace features in sections (veins)
1. Separate users
2. Workspaces
The research and ideas generated were further developed by the team of computer scientists and presented to the scientists to highlight new ways to explore their data and encourage them to continue to ideate and expand upon the tool. My role was designing the research plans, creating workshops to uncover the core functionalities, and synthesizing the data. From this work I was able to ideate and create a set of wireframes that I passed over to the other designer on my team. She polished the final interfaces while the computer scientist on our team developed the initial prototype. Currently the tool is in the hands of the research teams at Caltech to aid them in the process of sorting through their vast amounts of core sample data.
To understand the challenges of the researchers, we designed multiple sessions to observe their workflow and identify other groups of scientists who would be using the visualization tool. We did a “know” and “don’t know” exercise after our first interview to determine how to proceed with different methods and questions.
01. Researchers have no way to organize and keep track of their analysis of each individual core sample, or identify patterns between samples.
02. The limitation of current software restricts the amount of colors that can be mapped to each mineral, making finding relationships within mineral maps challening.
03. Scientists who are not familiar with spectroscopy and mineral maps are skeptical of the data and unsure on how to interpret it.
04. There are no ways to navigate through the 4,000 samples and understand “at a glance” what areas are interesting to investigate.
05. Sections were marked as “interesting” through an analysis done by hand, which is prone to errors and bias.
After gaining an understanding of the problem space, we had multiple design sessions with researchers to test out quick ideas and sketch new concepts together. This informed further ideation and the formation of resources to use in paper-prototyping.
We develoed low-fidelity paper prototypes to get a quick feel for the high-level functionalities needed for the tool. We tested the design with our two identified user groups (expert spectrocopists and scientists with no spectroscopy knowledge).
To understand the challenges of the researchers, we designed multiple sessions to observe their workflow and identify other groups of scientists who would be using the visualization tool. We did a “know” and “don’t know” exercise after our first interview to determine how to proceed with different methods and questions.
01. Researchers have no way to organize and keep track of their analysis of each individual core sample, or identify patterns between samples.
02. The limitation of current software restricts the amount of colors that can be mapped to each mineral, making finding relationships within mineral maps challening.
03. Scientists who are not familiar with spectroscopy and mineral maps are skeptical of the data and unsure on how to interpret it.
04. There are no ways to navigate through the 4,000 samples and understand “at a glance” what areas are interesting to investigate.
05. Sections were marked as “interesting” through an analysis done by hand, which is prone to errors and bias.
After gaining an understanding of the problem space, we had multiple design sessions with researchers to test out quick ideas and sketch new concepts together. This informed further ideation and the formation of resources to use in paper-prototyping.
We develoed low-fidelity paper prototypes to get a quick feel for the high-level functionalities needed for the tool. We tested the design with our two identified user groups (expert spectrocopists and scientists with no spectroscopy knowledge).
From our previous interview insights and information from the paper prototyping sessions we narrowed down on current, core, desired, and beyond scope functionalities for the design. I drafted documents detailing these functionalities in depth, and presented them to a group of stakeholders consisting of the core Caltech research team, NASA JPL HCI supervisors and mentors, the engineering team, and external reviewers. This set of functionalities was central to understanding the main features to prioritize in development, and which features the team could push to a future development cycle.
1. Display 3 minerals at once
2. Stitch together core sections (by hand or digitally)
3. Individually analyze and store notes
1. Display multiple minerals
2. Global and local spatial views
3. Analyze spectra
4. Filter and search based on target (mineral type, depth/section)
5. Mark uncertainties and discrepancies
1. Consolidation of core logs
2. Trace features in sections (veins)
1. Separate users
2. Workspaces
The research and ideas generated were further developed by the team of computer scientists and presented to the scientists to highlight new ways to explore their data and encourage them to continue to ideate and expand upon the tool. My role was designing the research plans, creating workshops to uncover the core functionalities, and synthesizing the data. From this work I was able to ideate and create a set of wireframes that I passed over to the other designer on my team. She polished the final interfaces while the computer scientist on our team developed the initial prototype. Currently the tool is in the hands of the research teams at Caltech to aid them in the process of sorting through their vast amounts of core sample data.
To understand the challenges of the researchers, we designed multiple sessions to observe their workflow and identify other groups of scientists who would be using the visualization tool. We did a “know” and “don’t know” exercise after our first interview to determine how to proceed with different methods and questions.
01. Researchers have no way to organize and keep track of their analysis of each individual core sample, or identify patterns between samples.
02. The limitation of current software restricts the amount of colors that can be mapped to each mineral, making finding relationships within mineral maps challening.
03. Scientists who are not familiar with spectroscopy and mineral maps are skeptical of the data and unsure on how to interpret it.
04. There are no ways to navigate through the 4,000 samples and understand “at a glance” what areas are interesting to investigate.
05. Sections were marked as “interesting” through an analysis done by hand, which is prone to errors and bias.
After gaining an understanding of the problem space, we had multiple design sessions with researchers to test out quick ideas and sketch new concepts together. This informed further ideation and the formation of resources to use in paper-prototyping.
We develoed low-fidelity paper prototypes to get a quick feel for the high-level functionalities needed for the tool. We tested the design with our two identified user groups (expert spectrocopists and scientists with no spectroscopy knowledge).
To understand the challenges of the researchers, we designed multiple sessions to observe their workflow and identify other groups of scientists who would be using the visualization tool. We did a “know” and “don’t know” exercise after our first interview to determine how to proceed with different methods and questions.
01. Researchers have no way to organize and keep track of their analysis of each individual core sample, or identify patterns between samples.
02. The limitation of current software restricts the amount of colors that can be mapped to each mineral, making finding relationships within mineral maps challening.
03. Scientists who are not familiar with spectroscopy and mineral maps are skeptical of the data and unsure on how to interpret it.
04. There are no ways to navigate through the 4,000 samples and understand “at a glance” what areas are interesting to investigate.
05. Sections were marked as “interesting” through an analysis done by hand, which is prone to errors and bias.
After gaining an understanding of the problem space, we had multiple design sessions with researchers to test out quick ideas and sketch new concepts together. This informed further ideation and the formation of resources to use in paper-prototyping.
We develoed low-fidelity paper prototypes to get a quick feel for the high-level functionalities needed for the tool. We tested the design with our two identified user groups (expert spectrocopists and scientists with no spectroscopy knowledge).
From our previous interview insights and information from the paper prototyping sessions we narrowed down on current, core, desired, and beyond scope functionalities for the design. I drafted documents detailing these functionalities in depth, and presented them to a group of stakeholders consisting of the core Caltech research team, NASA JPL HCI supervisors and mentors, the engineering team, and external reviewers. This set of functionalities was central to understanding the main features to prioritize in development, and which features the team could push to a future development cycle.
1. Display 3 minerals at once
2. Stitch together core sections (by hand or digitally)
3. Individually analyze and store notes
1. Display multiple minerals
2. Global and local spatial views
3. Analyze spectra
4. Filter and search based on target (mineral type, depth/section)
5. Mark uncertainties and discrepancies
1. Consolidation of core logs
2. Trace features in sections (veins)
1. Separate users
2. Workspaces
The research and ideas generated were further developed by the team of computer scientists and presented to the scientists to highlight new ways to explore their data and encourage them to continue to ideate and expand upon the tool. My role was designing the research plans, creating workshops to uncover the core functionalities, and synthesizing the data. From this work I was able to ideate and create a set of wireframes that I passed over to the other designer on my team. She polished the final interfaces while the computer scientist on our team developed the initial prototype. Currently the tool is in the hands of the research teams at Caltech to aid them in the process of sorting through their vast amounts of core sample data.
To understand the challenges of the researchers, we designed multiple sessions to observe their workflow and identify other groups of scientists who would be using the visualization tool. We did a “know” and “don’t know” exercise after our first interview to determine how to proceed with different methods and questions.
01. Researchers have no way to organize and keep track of their analysis of each individual core sample, or identify patterns between samples.
02. The limitation of current software restricts the amount of colors that can be mapped to each mineral, making finding relationships within mineral maps challening.
03. Scientists who are not familiar with spectroscopy and mineral maps are skeptical of the data and unsure on how to interpret it.
04. There are no ways to navigate through the 4,000 samples and understand “at a glance” what areas are interesting to investigate.
05. Sections were marked as “interesting” through an analysis done by hand, which is prone to errors and bias.
After gaining an understanding of the problem space, we had multiple design sessions with researchers to test out quick ideas and sketch new concepts together. This informed further ideation and the formation of resources to use in paper-prototyping.
We develoed low-fidelity paper prototypes to get a quick feel for the high-level functionalities needed for the tool. We tested the design with our two identified user groups (expert spectrocopists and scientists with no spectroscopy knowledge).
From our previous interview insights and information from the paper prototyping sessions we narrowed down on current, core, desired, and beyond scope functionalities for the design. I drafted documents detailing these functionalities in depth, and presented them to a group of stakeholders consisting of the core Caltech research team, NASA JPL HCI supervisors and mentors, the engineering team, and external reviewers. This set of functionalities was central to understanding the main features to prioritize in development, and which features the team could push to a future development cycle.
1. Display 3 minerals at once
2. Stitch together core sections (by hand or digitally)
3. Individually analyze and store notes
1. Display multiple minerals
2. Global and local spatial views
3. Analyze spectra
4. Filter and search based on target (mineral type, depth/section)
5. Mark uncertainties and discrepancies
1. Consolidation of core logs
2. Trace features in sections (veins)
1. Separate users
2. Workspaces
The research and ideas generated were further developed by the team of computer scientists and presented to the scientists to highlight new ways to explore their data and encourage them to continue to ideate and expand upon the tool. My role was designing the research plans, creating workshops to uncover the core functionalities, and synthesizing the data. From this work I was able to ideate and create a set of wireframes that I passed over to the other designer on my team. She polished the final interfaces while the computer scientist on our team developed the initial prototype. Currently the tool is in the hands of the research teams at Caltech to aid them in the process of sorting through their vast amounts of core sample data.