Trailblazer of the Week

Jihoon Yang

11 February 2022

The moment PCC student Jihoon Yang saw the announcement for the Lunar Trailblazer internship, he knew he had to get involved.

“My physics professor would make announcements to our class about interesting opportunities that are available to us students,” he says. “The Lunar Trailblazer internship immediately resonated with me, and after writing up my application and answering the technical questions, I went into the interview, and it launched forward from there!”

Yang is currently a PCC student hoping to transfer and pursue a degree in engineering physics. As a Science Data System (SDS) intern for Lunar Trailblazer, he writes Python scripts to build different computer vision and data processing methods for satellite image registration.

“I spend a lot of the time running different methodologies, comparing results, and hypothesizing the extent of the variables at play to improve future iterations,” says Yang. “Though there is theory involved, as well as learning from the works of others, a lot of it comes down to experimental trial and error.”

Yang’s project is part of Lunar Trailblazer’s novel approach to image registration for its data products. Most missions reconstruct where an image footprint is on the ground using only data on spacecraft position, orientation, and a camera optics model – basically trigonometry – to place the image at latitude-longitude coordinates on a planet’s surface. However, Trailblazer’s science requires much more precise localization of its data relative to existing images and topography of the Moon.

“Despite our Moon being our closest neighbor, there’s still so much we don’t know,” says Yang. “With water being such a key aspect for not just human life but almost all life as we know it, gaining a stronger understanding of the Moon by localizing water’s presence would be a major advancement!”

Working with SDS lead Co-Investigator and Caltech scientist Jay Dickson, Yang is helping create a workflow process that uses image-to-image tie points to locate future Trailblazer images more precisely. Lunar surface images are simulated from existing high-resolution lunar topography sun angle at time of Trailblazer observation. After an initial guess on spacecraft position, the algorithm scours the area in the simulated image for tie-points common between the simulated and actual data, for example, landforms like craters or scarps. With dozens of tie-points, the Trailblazer image can be iteratively localized to 1-2 pixel precision relative to the lunar surface.

For Yang, the excitement of being able to work in an environment where he can apply machine learning and computer vision provides a lot of motivation to learn new concepts and reinforce his skillset. This comes in handy when facing the real-time challenge of running experimental tests. When running new and data-rich image sets, computation times can be incredibly long, commonly lasting days.

“Whether or not our test scripts will yield informative results is not always a guarantee, meaning that there are weeks where tests may have been running but no new useful information is gathered,” Yang explains. “As a programmer, this is definitely a challenge that isn’t recognized often because computer science is usually an ‘instant results’ sort of process. However, it is a strong motivator to make sure that tests are run properly and efficiently, and gives me time to look over my own work and plan ahead for future iterations.”

Yang was born in South Korea but immigrated to the US and was raised here in Southern California. He’s always been fascinated by space, and as a kid loved reading books that deconstructed structural engineering projects, from medieval castles to modern machinery.

“I always found myself in awe of the engineering of the International Space Station and the Space Shuttle,” he recalls. “They both really made me fall in love with and respect the complex works that humanity can collaborate to build, as well as the amount of effort these large-scale projects require.”

Following his love for machines and aerospace, Yang went on to collaborate on flight simulator projects with classmates and become the Chief Technical Officer for his high school’s Air Force Junior Reserve Officers’ Training Corps. This would be the first of many significant mentorship roles he would assume.

“I helped build a flight simulator bay to teach students about the dynamics of flight and to support ground school operations,” says Yang. “Having learned programming at a young age, I’ve often found myself in mentorship positions for my peers, leading me to hold extracurricular academic sessions for students who wanted to learn well past our high school’s computer science curriculum.”

In turn, Yang’s mentorship experience has always helped rekindle his pursuit of a self-motivated education.

“To me, being in an environment where I can learn and be a part of novel experiences is the most important aspect of my aspirations,” he says. “They always provide a great amount of motivation back into my system to push me forward. That same drive continues to follow me working on Lunar Trailblazer!”

Even when outside the classroom Yang continues to find joy in building things, be it software or hardware.

“I spend a lot of my time working on personal projects and often enter into hackathons, which are competitions where people hack together cool projects in a finite amount of time,” says Yang. “Most recently, I attended HackMIT with a friend where we built a Chrome extension that’s able to determine whether or not text and images were generated by AI. We managed to win a prize for that submission! More recently, I've been learning how to work with circuitry, so I hope to use those new skills at a hackathon soon to build even more interesting projects.”

As Yang continues his SDS internship with Lunar Trailblazer, he reflects on his first experience working on a space mission, and ponders how those pesky computational wait times aren’t the only aspects of his role that require patience and trust.

“I really can’t put into words the happiness of being a part of this mission,” he says. “It can definitely be difficult at times to judge progress in the moment. It can feel as though you're stagnant and not moving forward. However, being able to trust yourself and the fact that your efforts are there, turning some gears, is incredibly important. Believe in yourself and work hard, even if it feels as though things are working against you, so that one day you can look back and see the trail you blazed—haha!”

Jihoon Yang is a PCC Science Data System intern and Trailblazer of the Week!

Trailblazer of the Week is an ongoing series showcasing the diversity of experience and expertise that supports the collective determination of the Lunar Trailblazer mission.

By Emily Felder
Emily Felder is a Pasadena City College student and Caltech intern working on science communication for the Lunar Trailblazer mission.