Project 7 of 7

NASA SOFIA Internship

A summer at SOFIA, the airborne infrared observatory, and the problem I went looking for

Overview

Finding a Problem Worth Solving at 45,000 Feet

In the summer of 2013, through the California State University’s STAR program, I was placed at SOFIA, the Stratospheric Observatory for Infrared Astronomy, based at NASA’s aircraft facility in Palmdale, California (STAR Program; Cal Poly CESAME; Wilson et al., 2013).

SOFIA was a Boeing 747SP modified to carry a 2.7-meter infrared telescope, with an effective aperture of 2.5 meters. It flew at up to 45,000 feet, above most of the water vapor that blocks infrared light from reaching the ground. It was a joint program of NASA and the German Aerospace Center (DLR), with science operations coordinated by USRA in the U.S. and the Deutsches SOFIA Institut in Germany. Flight operations ended in September 2022 (Rosson, 2025; NASA, SOFIA; DLR, 2022; Deutsches SOFIA Institut).

It was an extraordinary place to spend a summer. I didn’t have a project.

SOFIA, a modified Boeing 747SP, in the hangar at Palmdale.
SOFIA in Palmdale, summer 2013. [CAPTION — WHAT AND WHEN]

My mentor was based at another location, so when I arrived the work hadn’t been defined yet. I started asking people around SOFIA a simple question: is there anything here that needs to be figured out?

An electrical engineer showed me a system that had been producing strange readings. That became my project.

Method

Finding the Anomaly

The data came from sensors on SOFIA’s Telescope Assembly. Earlier analysis had looked at one flight at a time. I compared fourteen flights side by side, looking for behavior that held from flight to flight and for the flights where it didn’t. [VERIFY: fourteen flights — the summary chart’s axis runs 96 to 107]

The system I focused on was the Power Bus System, which monitors the current [VERIFY: current, voltage or load] across the three phases of each of three power buses feeding critical onboard electrical systems. In a balanced three-phase system the phases run offset by 120 degrees, so the difference between them should stay small. A growing difference is a sign something is wrong.

The work grew into [VERIFY: several thousand lines of R code], along with plots, diagnostic tools and data files for later analysis.

The question shifted from whether something unusual had happened to what abnormal behavior looks like, consistently enough that software could recognize it.

Results

Normal, a Spike, and a Sustained Problem

Comparing flights made it possible to tell three kinds of behavior apart on power bus 3.

Flight 101: normal

Power bus 3 on flight 101: difference between phases stays below the threshold line for the whole flight.
Flight 101, power bus 3. A normal flight. The difference between the three phases stays below the proposed threshold, the dotted line, from start to finish.

Flight 103: a spike

Power bus 3 on flight 103: a spike above the threshold early in the flight.
Flight 103, power bus 3. One of two flights with an abnormally high maximum average difference. Early in the flight the difference between the phases spikes above the recommended threshold.

Flight 104: sustained

Power bus 3 on flight 104: difference stays above the threshold for nearly the whole flight.
Flight 104, power bus 3. The difference from the rolling average stays above the threshold for nearly the entire flight. The same problem appeared alongside other electrical-system issues on that flight.
Maximum average difference for each of the three power buses across the flights, with flights above the threshold in red.
Maximum average difference, all three power buses. Flights above the suggested threshold are shown in red. A closer look at the anomaly on power bus 1 showed it was instantaneous, not sustained.

That distinction is what the threshold is for. A single spike and a sustained imbalance call for different responses. The analysis set a threshold that could be programmed into warning software to alert operators to power-load problems. [VERIFY: proposed or implemented; the threshold value and its units]

I applied the same multi-flight approach to gyroscopic drift, looking for trends that would help future flights track more reliably. [VERIFY: what was found, or cut this paragraph]

Nobody handed me a canned exercise. I went looking for a problem, found one the people working on the aircraft cared about, and built the analysis around it.

Missions

Flying on SOFIA

I flew on two SOFIA missions, on July 2 and July 8, 2013. These were overnight observing flights aboard a working observatory, with scientists and engineers running the telescope and instruments while the aircraft flew its observing route. [VERIFY: targets observed — Jupiter, Galactic Center]

Faces of other crew members are obscured for their privacy.

Steven Wilson aboard SOFIA during an observing mission.
Aboard SOFIA, July 2013. [CAPTION — DATE AND WHAT IS SHOWN]

SOFIA was loud. It was a 747 rebuilt around a telescope, not a passenger plane with one bolted into the back, and talking aboard meant hearing protection and headsets.

Mission workstations inside SOFIA during an observing flight.
[CAPTION — WORKSTATION / INTERIOR, DATE]
A yellow crocheted octopus with black eyes and a pink smile.
My internship mascot, summer 2013.

Field visits

Beyond the Analysis

Inside the telescope

I got much closer to the telescope than any airline passenger would: booties, hair cover and lab coat, and into the area around the telescope assembly itself. The infrared instruments need cryogenic cooling, so the internship included cryogenic-safety training, a reminder that astronomy sometimes means learning how not to asphyxiate yourself. [VERIFY: cryogen details]

The SOFIA telescope area, with protective clothing.
[CAPTION — TELESCOPE AREA, DATE]

A large mirror needs a large workshop

I also toured the equipment used to maintain SOFIA’s mirror, including the vacuum chamber used in the coating process. I didn’t see a coating run, but the machinery gave a sense of the scale behind a telescope usually described in a single number.

The vacuum chamber used to coat SOFIA’s telescope mirror.
[CAPTION — MIRROR COATING CHAMBER]

A side trip to Edwards

The internship included a visit to Edwards Air Force Base, including a small rocket-engine fabrication and machining shop and video of engine testing. Then there was the anechoic chamber: phones stayed outside, and every surface was covered in the pyramidal absorbers used for electromagnetic testing. [VERIFY: facility name and size claim before naming it]

Edwards Air Force Base.
[CAPTION — EDWARDS, PUBLICLY USABLE PHOTO ONLY]

Outcome

From No Project to a Research Poster

What started with no defined project ended as a multi-flight analysis of real SOFIA engineering data, presented as a research poster. [VERIFY: where and when presented]

Tools and Methods to Optimize the Analysis of Telescopic Performance Metrics on SOFIA

Steven R Wilson, Holger Jakob, Stefan Teufel, Zaheer A Ali, Jeffrey Van Cleve, Brian Eney, Greg Perryman. CSU Fullerton, Deutsches SOFIA Institut, USRA (Wilson et al., 2013).

The complete research poster, Tools and Methods to Optimize the Analysis of Telescopic Performance Metrics on SOFIA.
The complete poster. The fourteen-flight method, the power bus analysis, the proposed warning threshold, and the gyroscopic drift work.

Sources

References

  1. Cal Poly Center for Engineering, Science and Mathematics Education (CESAME). About CESAME. California Polytechnic State University, San Luis Obispo. cesame.calpoly.edu
  2. German Aerospace Center (DLR) (2022, April 28). The airborne observatory SOFIA to end its mission after eight years of operations. Joint statement by NASA and DLR. dlr.de
  3. Deutsches SOFIA Institut, University of Stuttgart. Institute. dsi.uni-stuttgart.de
  4. NASA. SOFIA: Stratospheric Observatory for Infrared Astronomy. blogs.nasa.gov
  5. Rosson, L. R. (2025). SOFIA: A NASA History Office Report (NASA SP-2025-4901). NASA History Office. nasa.gov
  6. STAR (STEM Teacher and Researcher) Program. STAR Program research presentations. Digital Commons @ Cal Poly. digitalcommons.calpoly.edu; starteacherresearcher.org
  7. Wilson, S. R., Jakob, H., Teufel, S., Ali, Z. A., Van Cleve, J., Eney, B., and Perryman, G. (2013). Tools and Methods to Optimize the Analysis of Telescopic Performance Metrics on SOFIA. STAR Program research poster. Lab site: NASA Armstrong (formerly Dryden) Flight Research Center. digitalcommons.calpoly.edu