NASA summer intern Grace McFassel in her lab. Credit: NASA/Emily Cavanaugh
SCaN intern project, optical communications 

An Odyssey at the Oscilloscope: NASA Intern Automates Optical Signal Processing

By Emily Cavanagh

August 15, 2019

It’s 4 p.m. on a Friday, and Grace McFassel, a Space Communications and Navigation intern at Goddard Space Flight Center in Greenbelt, Maryland, has barely been to her office all week. Instead, she has been huddled over her lab bench wearing a blue coat. She’s tethered to the bench by an ESD anti-static wrist strap, an essential accessory for anyone working closely with delicate electronics. As you’ve probably guessed by now, hers is no ordinary summer internship.

“You see these two bars?” she asks, pointing to an oscilloscope screen. “These are two of our signals.” The hum of technology radiates through the lab as McFassel explains the goal of her project, which involves aligning signals in order to improve the performance of modems. “If you don’t have these lined up with each other at small tolerances, you’re not going to be able to run at the speeds you need to run at.”

In optical communications, lasers beamed through space provide unprecedented communications data rates, but without modems that can process and encode data at these speeds, the advantages of optical communications are moot.

The control signals for this optical modems are divided into eight waveforms. They have to be exactly in sync to achieve their desired data rate – 10 Gigabits per second, or about 250 typical MP3 files per second. This alignment allows Field Programmable Gate Arrays (FPGAs), programmable semiconductors used in computing, to handle the high-speed data transfer.

Alignment was previously done by hand, but this summer, McFassel wrote code and built simulated environments, or test benches, that could automate tap alignment for FPGAs. McFassel ran tests in these environments and examined the timing results, optimizing the system as the summer progressed. While her work focused on tap alignment for 10G modems, it could be applied to any project where multiple lines of data must be synchronized.

For McFassel, whose doctoral work at Texas A&M University centers on sensor theory for robotics, FPGA programming wasn’t a new challenge. “It was an interesting opportunity to take skills I learned in undergraduate, keep them fresh, and use them for something practical,” she said.

For McFassel, much of this summer was as about viewing familiar territory through a new lens. She grew up in the small town of Milford, Delaware, and earned her undergraduate degree in computer engineering from the nearby George Washington University in Washington, a quick 30-minute drive from Goddard. Her older sister, Rose attended the university at the same time.

“We're very close and she's always been really supportive,” McFassel said of her sister, who she credits for inspiring her to persevere in her major. “Being an engineering student felt so hard and I almost felt like I couldn’t do it. She'd metaphorically whap me with a piece of paper – sometimes literally. She always told me to never give up."

Another inspiration in her work is Rick Butler, McFassel’s mentor this summer. Butler called McFassel “a quick study,” adding that he liked her enthusiasm and curiosity about the work they did together.

“We gave her a fairly complex project, and she was able to learn a new set of software tools in fairly short order,” Butler said. “This allowed her to generate an initial design and deploy hardware in a few short weeks.”

McFassel’s interest in computer hardware can be traced back to her childhood.

“One of my earliest memories is of my family actually getting our first desktop computer in the house,” McFassel said. “We had educational games, and I'd see them and go ‘How does that work? What is that? I see the pictures, but what’s actually going on?’"

McFassel started building her own computers in high school, eventually using parts from that first desktop computer. Now, McFassel’s research into sensor design concerns what information is needed to build autonomous robots.

“My interest is in understanding what sensory information is actually used, and what part of that information is just extra,” McFassel explained.  “By doing that, we're hoping we can help designers make smarter systems.”

The systems that could benefit from her research range from self-driving cars to robot vacuum cleaners.

This fall, McFassel will begin a fellowship with 3M, a multinational conglomerate focused on science and innovation. The company is researching sensor design for automated vehicles. She plans to integrate her research with theirs in the coming year.

Outside of robotics, McFassel has multiple creative outlets. She enjoys baking, sewing and playing video games.

“I learned how to sew partially from my mom, partially from the internet, partially from lots of trial and error, but I love to sew dresses and costumes,” said McFassel. Together, she and her sister have previously designed costumes freelance.

As the summer of 2019 came to a close and McFassel headed back to Texas to continue her doctoral work, she offered the following advice for others thinking of pursuing their dreams at NASA or elsewhere:

“I'm a big proponent of that do what you love, follow your dreams kind of stuff,” she said. “My interest in computers sparked from a very early age from just wanting to know more about them, and I always chased that. So, follow your passions — no matter what.”