This is just the very best thing, possibly ever.
Among the several fun things about this Twitter adventure has been the occasional interactions with ACTUAL NASA PEOPLE! Real engineers and programmers and scientists who actually do all the things that I then turn around and make fun of like a jerk.
So, in what I hope will be the first of many such happenings, I reached out to one of these people in the hopes of asking some questions… and as a result we have today’s post.
An interview with MARS ROVER DRIVER MATT HEVERLY! (AKA @Matt_Heverly on Twitter)
So click through to learn “All I Want To Know About Rover-Driving, But Was Unable To Ask*”
This interview has been broken into two parts for length and in order to annoy you…
SR: Matt, thank you so much for agreeing to this. Hopefully I don’t ask the dumbest questions in history. Preliminaries first: What is your name, location, age, job, and level of awesomeness?
MH: Matt Heverly, Live and work in Pasadena, CA. 36 years old (born on leap year so actually only 9). I worked at the mobility systems engineer for MSL for 2 1/2 years and am now the lead rover driver. There are 16 rover planners who are responsible for driving the rover, operating its robotic arm, and the turret instruments at the end of the arm. Each of us has a different expertise, some the arm, some the simulation software, and me mobility. Brian Cooper is our rover planner team lead and is super awesome. He was a rover planner for Sojourner, both MER rovers and MSL. He is the team lead, but I am in charge of mobility which makes me the “lead driver”. I am going through a lot of detail to explain this because sometimes people get it wrong and say that I am the team lead and I don’t want to take anything away from Brian who has an awesomeness level of 10. My awesomeness level is much lower.
SR: In as much detail as you like, how did you become a Mars-Rover Driver?
MH: I literally received and email in March of 2005 asking if anyone was interested in MER operations. At that point the project had been going on for a whole year and people thought it was only going to last three months. People had other commitments and started leaving the project and they needed to backfill the staff. I started my career at the company that build the robotic arms for the MER rovers. I then went to graduate school and after school started working at JPL. When they asked for volunteers for MER I jumped and since I had experience with the hardware and a robotics background I got the gig. I trained for about year and became a rover planner for Opportunity. I worked on MER half time until I left to work on MSL in early 2010. For the last two and a half years I have been living MSL mobility. I have been working with the people who built the hardware, people who wrote the software, testing the system, and driving the twin earthbound rover in the Marsyard nearly every day getting ready for the chance to do it for real on Mars.
As an aside, while working half time on MER half time I spent my other half time building a big robot on roller skates called ATHLETE.
SR: What is the process for driving a rover on Mars? Is there a joy stick?
MH: Most people picture us with an x-box joystick driving the rover around Mars. The problem with that is the time delay. If we saw a giant cliff in front of the rover, by the time we hit the stop button it would be 14 minutes before that command got to the rover. Instead of using a joystick we give the rover a list of “tasks”. We talk to the rovers roughly twice a day. In the (Mars) morning we send it a list of things we want it to do that day: drive forward 1m, turn right 10 degrees, take some pictures, etc. At roughly 4:00pm Mars time the rover “emails” us the results of what it did that day (telemetry and pictures). The rover then goes to sleep since it is really cold at night on Mars and takes a lot of power to keep it warm. While the rover sleeps we build sequences for the next day. We then repeat that cycle every day. This is why the people working on the project have to live on Mars time. We always come to work at essentially 4:00pm Mars time, independent of what time it is here on Earth.
SR: What kind of computer interface are you working with? Are you driving in 3-D, or just inputting text commands and waiting a day for results?
MH: We have a custom simulation call the Rover Simulation and Visualization Program (RSVP). Brian Cooper who I mentioned before and has an awesomeness level of 10 is the lead for that project as well. You can see an animation of our simulation at the link below.
This simulation is in 3D and we use 3D glasses to view the terrain and rover. It seems like this might just be a novelty but seeing the terrain in 3D makes a big difference in being able to identify hazards.
SR: Is the rover mapping its surroundings to help you out?
MH: The rover has the ability to make decisions on its own. It can image the terrain using stereo cameras and evaluate whether the terrain is safe or not. If there is a big rock in front of the rover and we tell it to drive straight, it will stop and call home for help .
SR: Who decides what path to take? Is the rover at all autonomous once it’s begun a drive?
MH: We typically drive “blind” the first 20 or 30 meters. This is as far as we can really see reliably with the Navcams. Once we go beyond the Navcam horizon we turn on autonomous mode where the rover keeps itself safe. The reason we do this is because it is about 3x faster in blind mode where we tell the rover “trust us, there are no rocks in front of you”. This saves the time of imaging and evaluating the terrain and with the relatively slow processor on the rover it saves a lot of time. The fastest we can ever drive, just spinning the wheels as fast as they go is 151 meters/hr. When we drive in autonomous navigation mode it is 40 meters/hr.
SR: If you were to think of Curiosity as having a personality, how would you describe her?
MH: She kicks ass. Every test that we throw at her she always outperforms expectations.
END OF PART ONE:
Hi Jason,One other interesting thing that I was thinking about last night is why the rover drivers get all the press. I don’t think we have the hardest job, that belongs to the Tactical Uplink Lead (TUL).The TUL has to know everything about the rover. They manage the science team and all of their observations. They manage power, data volume, communication passes, etc. But when you talk to someone who isn’t involved in the project and tell them you are a TUL they say “huh”? When you tell someone that you are a rover driver though they have an idea of what you do…it is typically wrong since they picture us with a joystick, but I think that our job is the most relatable for the general public.The entire team is full of amazingly talented people and I feel a bit of guilt that the rover drivers get all the press, but I also think it is important to engage the public since the rover really belongs to everyone. All the raw images are posted on the web almost immediately after they are returned to Earth. All the scientific data is published publicly and we are funded with tax dollars allocated to NASA (the American public funded this mission).I think we have an obligation to let the public know what we are up to, but I want to make sure the entire team gets the props that they deserve.-Matt