Talos

A diagnostic tool for NASA that enables astronauts to autonomously troubleshoot and fix anomalies in deep space.
Role

Experience Design
Research
Strategy

Company

NASA
(Capstone at Carnegie Mellon University)

Duration

8 months (2020)

Team

Katie McTigue, JT Aceron
Megan Parisi, Nathan Barnhart

PROBLEM

Astronauts rely heavily on Mission Control (a 40+ personnel ground team) on earth to solve anomalies.

However, a mission to Mars means long communication delays (almost 44 mins) and blackouts, which means astronauts will have to work more autonomously from Mission Control while diagnosing and troubleshooting problems.

RESEARCH

To understand how astronauts diagnose problems and the future of troubleshooting we looked at 4 different things:

1. Background research

(4 research professors, 20 research papers, 3 commercial space employees)

Our research helped us identify some of the key stakeholders. Mission Control or MCC consists of Flight Controllers who are subject matter experts on particular submodules of the International Space Station (ISS). Today ISS is almost entirely controlled by MCC.

2. Johnson Space Center

(9 interviews with Flight controllers, Astronaut Trainers and other experts)
We were incredibly lucky to go to the Johnson Space Center in Houston, the home of Mission Control! We sat in on a flight controller training session for some contextual inquiry, saw Mission Control, went inside a full scale Mockup of the International Space Station (ISS) and talked to subject matter experts.

Here’s what we learned:

1. Flight controllers “special skill” is real-time quick decision making.

“Nothing can outmatch the human brain when it comes to critical thinking.” - Gateway Operations Engineer

2. Situational awareness is fluid- The crew with their physical presence or MCC with massive amounts of data can have more situational awareness depending on the situation.

“In high stress situations, the crew has more situational awareness, and the ground does not provide good judgement.” - Flight Controller

3. Astronaut Perspective

(2 astronauts, 2 astronaut books and 1 Extra Vehicular Activity)
Getting in touch with astronauts is as hard as it sounds, but I was determined to talk to our primary user. After sending out multiple cold LinkedIn messages, two astronauts agreed to talk to us.

Here’s what we learned:

1. Slow is fast- Taking time diagnosing anomalies reduces mistakes.

“Rarely a time when anomalies require you to act quickly and make things worse.” - Astronaut on Soyuz-10 in 2018

2. Procedures are strongly ingrained and crew follows them most of the time.

“Through training, there are pre thought out decisions when something goes wrong.” - Astronaut on Space Shuttle Atlantis in 1985

3. Psychological strain greatly impacts crew.

“Trying to fight the monotony / psychological welfare will probably become driver in how you schedule the crew.” - Astronaut on Soyuz-10 in 2018

The process we used to arrive at findings presented above- Creating an affinity map.

4. Analogous domains

(5 interviews + online surveys)
We wanted to see how experts diagnose anomalies in other problem spaces, so we turned to Analogous domains. We interviewed HVAC technicians, miners, electrical and computer engineers and sent out online surveys to auto mechanics and pilots. We realized two things:

There is temptation to cut corners during the diagnostic process.

“Even the most experienced engineers break the rules sometimes.” - Electrical Engineer

It is important to track and document the diagnostic process.

“Diagnosis is a backwards process. You’re starting with the end result, the problem, and you need to backtrack using the schematic to see where in the process something has gone wrong.” - HVAC Technician

BUILDING EMPATHY

We built empathy with astronauts understanding how communication delay and psychological effects impact their diagnostic process. We did this with an empathy building game and diary study.

Empathy Building Game

We played a cooperative puzzle game called Keep talking and nobody explodes where players have to diffuse a bomb. We added a 10 second time delay to the game to simulate the delay astronauts might experience. Mistakes had more weight as delays increased and people naturally tended to front-load information to combat delays.

Diary Study

Covid-19 crisis changed our world, but it presented an opportunity for research. We conducted a quarantine diary study to understand the effects of isolation on people, as a way to understand the psychological effects astronauts face in space. We surveyed 48 participants over a 10 day period. We learnt that food is often the high or low point of a persons day. Days start blending together and it becomes hard to distinguish special days.

EVALUATING NEEDS

Through our research, a lot of astronaut needs started to emerge. We tested these needs by creating storyboards and presenting them to astronauts! Here are some examples of our storyboards:

How might we keep the crew going when things get tough? Need Validated

How might we condense information into a single screen? Need Validated

How might we make safety and engineering data more available for just-in-time training? Need Validated

How might we optimize diagnostic schematics? Need Validated

How might we make it easier for astronauts to follow multiple procedures? Need Validated

How might we manage conversations with long delays? Need not validated

How might we help people practice troubleshooting best practices even when they get frustrated and suffer from the “space stupids?” Need Validated

How might we help astronauts remember complex systems training? Need Validated

How might we update the concept of procedures for unanticipated anomalies? Need Validated

How might we make authority clear? Need not validated

How might we help astronauts characterize risk? Need not validated

This process worked really well, because astronauts were very honest. They told us when a need was valid, and when something wasn’t our strongest idea.

KEY INSIGHTS

From all our research here are our key insights:

1. Previous experience, stress and physical effects can cause crew to stray from best practices.
2. Current system relies heavily on ground and following procedures, which limits crew’s opportunity to think critically.

3. The crew and the ground have different mental models.
4. Reference materials during diagnosis are just as important as pre-flight training

NARROWED PROBLEM SPACE

Based on our insights and a reframing exercise, we crafted a narrowed problem statement:

How might we situate diagnostic tools in a workflow to support crew members in thinking clearly, rationally, and critically?

We see three core concepts in this statement- Documentation, Best Practices and Critical Thinking.

So far we have explored a vast number of research avenues before synthesizing our findings into a final problem to solve- supplementing a diagnostic workflow. In the summer we will make this solution space a reality through ideation, user testing and final product development. So stay tuned!

KEY TAKEAWAYS

Being resourceful (interviews with astronauts): Initially we were told by the client and faculty that no team had managed to talk to an astronaut. However, I was very determined to talk to my end-user, so I reached out to multiple astronauts on LinkedIn and email. Persistence and luck led to me to find 2 astronauts and positive feedback from them during our presentation helped us build trust with our clients. They then connected us to 3 more astronauts!

Dealing with differing opinions (astronauts vs MCC): Astronauts and MCC often had differing opinions- Level of autonomy, Features required, etc. Since our end user is the astronaut, I placed more weight on their opinion while keeping SMEs views from MCC in mind.

Research in unanticipated circumstances (Covid-19): Covid-19 completely changed the way our team worked- everything had to be moved online, so we used tools like Miro. The way we researched had to change as well, but we used the online platform and social isolation to our advantage- To build empathy with end users by conducting a study on isolation.

Link to our team blog:

NASA X MHCI