AD (728x60)

Feature Post

Saturday, January 17, 2015

Tetramorph Avionics: My Experience Building a Lunar Rover

Share & Comment
Team Photo From End of Semester Presentation
In most courses I have taken throughout my life, the closest that I have gotten to a real-world experience is reading about it in a textbook.  However, this semester, I had the chance to take a course that truly immersed me in a real-world experience, teaching me many lessons both technical and social and giving me the once-in-a-lifetime opportunity to work on one of the coolest things out of this world: a lunar rover.

An Inspiring Video about Andy created by Luis Ball and the Lunar Rover Team

Carnegie Mellon’s hidden gem, 16-861, Mobile Robot Design (or 16-865 in the spring) was truly a life-changing experience.  Going in, I expected to work on Carnegie Mellon’s prized lunar rover, Andy.  However, this all changed when Professor Red Whittaker, the head of the lunar rover team and chairman of Astrobotic, unveiled a new project possibility.

This new rover would go by the name of Tetramorph.  The vision was to create a portable and cheap rover that could be used for high-risk missions, without the worry of destroying a mission critical rover.  The idea makes sense, as it would be detrimental if Curiosity or some other national figure was taken on high-risk missions and destroyed.


The USSR Lunokhod - The Rover That Started It All
By Clemens Vasters from Viersen, Germany, Germany (USSR Lunokhod 1  Uploaded by russavia) [CC BY 2.0 (http://creativecommons.org/licenses/by/2.0)], via Wikimedia Commons
The course began by preparing all of the space newbies (including myself) by giving a brief history of rovers and teaching us about many of the challenges they faced.  Then, we were assigned a temporary project to research and present a proposal of semester goals; I was assigned Tetramorph computing.  Little did I know that Tetramorph computing would become a large part of my life over the next semester.  


Carnegie Mellon Professor Red Whittaker loves pushing missions to the borderline of impossible.
Image via lunar.cs.cmu.edu
After the goal proposals, everyone in the room was given the chance to divide into two teams: the new mysterious Tetramorph and the already prominent Andy Rover, Carnegie Mellon’s Lunar X Prize treasure.  It was a difficult decision to make; should I join an already established team that is bound to go somewhere, or should I risk giving up a semester to work on a project that could fail and be scrapped?  Professor Whittaker made it clear that the Tetramorph was at the borderline of impossible, so there were no surprises associated with the risk.


For those new to space and aeronautics, the word avionics can be full of mystery.
Image via Google
Ultimately, I had decided, along with around eight others, to sign up for the adventure that was Tetramorph Avionics.  We jumped right into work, looking into how we would design the system and the various components we would be using.  However, we faced an issue much larger than that of finding the appropriate components; every week, someone would disappear.  Whether it was an excuse in a Facebook message or a sincere apology, people began to leave as they saw the large time commitment required to make this thing run.

Finally, we ended up with a team of four undergraduates: Abhinand, Edwin, Jaime, and I.  While most teams had several master students and grizzled veterans, Tetramorph Avionics consisted of a measly two juniors and two sophomores.  Could we do it?  Would we get something running by the end of the semester?


A special guest during our presentation in hopes of recruiting new members.
After we realized just how serious this issue of team size and experience had become, we found ourselves looking around intently for new Tetramorph Avionics members.  It even involved me unveiling an Uncle Sam slide during a presentation in hopes of recruiting new members.  It turns out that not many people know about electrical systems, and those that did were already fully committed to the Andy Rover team.  This slide did little more than add a bit of humor to our slides and make our confidence look diminished in front of the rest of the team.

Realizing that we were out of options for recruiting new members, we knew that the only way we would achieve our goal of creating a Tetramorph electrical system and getting it running was if we moved forward with full force.  We quickly became more organized, dividing and conquering tasks and preparing our first parts list.


Our First Big Parts List Order
I can still remember the night that we sent in our first parts list like it was yesterday.  After someone mentioned the idea of getting the parts list ready another day, I looked at them and made it clear that I was not leaving until I had a list of parts to send to Joe, the class TA.  Determined, we got that parts list in.


The Team That Made It All Happen
In 16-861, each team presents a weekly progress update.  This was exciting and interesting, as it allowed everyone in the class, around 40 people in total, to keep up with what the rest of the teams had accomplished.  This kind of team presentation had built up the kind of team momentum like those found in a NFL football team.

Throughout the semester, we quickly went from the team that had practically begged for new members to a team that objectively presented the goals it had achieved and the goals for next week in a professional and impressive manner.  Our presentations got more informative, and we continued to grow as a team.

The Tetramorph Rover with Andy at its Side
In addition to presenting to the class as a whole, we had weekly Tetramorph meetings.  These were ran by Jay, mechanical guru and visionary of the folding Tetramorph design and Curtis, who was raising both a rover and a child at the same time.  Both were determined to have a finished product by the end of the semester, and both had aided us greatly in giving advice and helping when needed.  They, along with everyone on the Tetramorph team, put themselves entirely into this project.


When we got the motor controller working, it responded with a smile.
As progress was made, the semester, began to speed up.  Parts lists continued to be submitted; parts continued to arrive.  We began drafting schematics, breadboarding circuits, and discussing plans for the Tetramorph as a whole.  Incredible accomplishments had begun to appear all around us.  Tetramorph Mechanical made its first wheel.  Tetramorph Software made its first point cloud.  Video guru Luis released another Andy video that inspired us all.  

The Exciting Unveiling of Andy
Image via lunar.cs.cmu.edu
Fast forward to November and we found ourselves part of a formal unveiling of the Andy Rover.  It was a release mirroring that of a company like Apple, with the Andy Rover being driven out on a robot table, complete with a smoke machine and LED lighting effects.  Andy was even spray painted gold to look pretty for the press.  The event was well-received by the press and made its way on many news sites and local news channels.


Jay (right) with the Folded Up Tetramorph
As the semester neared its end, things sped into overdrive, as the whole of Tetramorph was determined to produce a functioning rover.  One day, Jay came in with a fully assembled mechanical Tetramorph rover.  It was one of the most beautiful things I had seen in my life; it was clear that the number of hours that he and the rest of mechanical had placed into this was uncountably high.  Software had the same kind of crazy progress, as they had created a fully testable system on a push cart that mapped out its environment and told them where to move the cart for testing.  

Now, the tension was on; as the Tetramorph Avionics team, we did not want to let the rest of our team down.  We wanted to strike the rest of the team with awe and do the impossible: make the rover move by the end of the semester.  A week before the end of the semester, when I mentioned a moving rover to Professor Whittaker during my one-on-one chat with him and Joe, he became a bit surprised at this statement.  With the little time remaining in the semester, this seemed like a difficult goal.


A Breadboarded RS-485 Network
During the last week of the semester, we spent late nights working on a multi-node communication network, developing a power system, and fitting the electrical components into a protoboard.  Our first issue arose when we tried to connect half-duplex RS-485 motor controllers (two wire protocol) to a full duplex RS-485 network (four wire protocol).  After much attempting and failing, we had realized that this would just not work as easily as expected.  

Fast forward to 5:00 pm the night before the final presentation, and we faced another issue: the power system was not ready to test the wheel.  After a bit of confusion, tension rose between members of the Tetramorph Avionics team that threatened the success of the Tetramorph rover.  However, with the help of Curtis, we were quickly able to get our heads back in the game and continue making progress. 


A Tetramorph Wheel
Finally the power system had been complete.  As time drew near to the end, we realized that we were not going to get this network working with the motor controllers.  When I asked Jay about simply connecting the motor controllers to the main computer via USB temporarily, he thought the idea of fitting USB cables into the small wheels was crazy.  I realized how crazy this was and began carving away at the end of a USB cable in hopes of making it smaller.  It turns out there is a lot of rubber at the end of a USB cable, and the metal end itself was not that large.  People thought I was crazy as I shredded into a USB cable mercilessly.

We filled a wheel with the electrical components and flipped the switch.  All lights had turned on as they should have.  We then plugged the motor controller into our computer and sent the command: the wheel began to spin.  Following this success, we then connected the motor controller to the main computer of our rover and remotely connected to it.  We were able to send the command over the local network to make the wheel spin using a C program; this was one of the most exciting moments of the semester.

Getting the Tetramorph to move (Jay on left, Edwin on right)
Video via Tetramorph Final Presentation

We quickly duplicated another wheel protoboard and installed it in the wheel.  However, this time, something was wrong: the motor controller would not turn.  We spent late into the night attempting to get this working.  At around 2:30 AM, I left, so that I would be able to prepare for the big presentation the next day; I would come back in the morning to see if I could get it running.  However, Edwin and Jay had the perseverance to stay with that rover until 4:00 AM.  They fixed the issue with the second motor controller and successfully demoed the Tetramorph running.  They had gotten it to move successfully and even run over an obstacle.


The Lunar Team's Amazing New Website (lunar.cs.cmu.edu)
The next day, we had our big presentation.  It opened up with one of Luis’s inspiring videos and each team rose to the front to explain the amazing amount of progress they had made.  In one semester, the Andy Rover got a large amount of updates, had a huge unveiling, and won the $500,000 Google Lunar X Prize Milestone Award.  In one semester, a collection of professional videos had emerged and an awe-inspiring website (lunar.cs.cmu.edu) was created.  In one semester, we had gotten the Tetramorph moving.

Reflecting back on this project, it is hard to believe the sheer number of people that put themselves entirely into this project.  They truly dreamed of bringing Carnegie Mellon to the moon.  Focusing in on the Tetramorph team, those involved in the mechanical and software teams had done some amazing things.  A full mechanical design of a rover and extensive software tests would probably take ten times as long in the existing space industry atmosphere.  And finally, looking back at the Tetramorph Avionics team that I was part of, I cannot believe what we had accomplished.

Abhinand, Edwin, Jaime and I had thrown ourselves into the project and come out successful.  We hit a few road bumps along the way and the path was far from smooth, but we had remained together and persevered through it all.  While we had differences at times and the road was not always clear, we knew that we needed each other more than ever and if one of us left, the Tetramorph Avionics team would fail.

Our experience this semester was one like no other.  If you are looking for a course that gives you a once-in-a-lifetime experience, I highly recommend this course.  If you truly become committed, that which you will take away from this course is beyond your imagination.



Tags: , , , , , , , , , , , , , ,

3 comments:

  1. Wow! After reading your blog post, I could not be more excited to join my school robotics team! I am currently a sophomore at Eleanor Muskoque High School. I hope to someday apply to Carnegie Mellon University. Our school robotics team competes in the FIRST Robotics Competition and we take it very seriously. Our school was the FIRST Champion last year for our region. Did you participate in FIRST Robotics back in high school? Would you say that the experience you received from your FIRST experience has helped you build the Tetramorph? Or if not, do you believe that having FIRST experience is not important to college robotics and has not affected your current lunar rover project in a meaningful way (you learn how to build a fully functional robot from scratch in FIRST)?

    ReplyDelete
    Replies
    1. That's awesome to hear that you are getting involved in robotics already in high school! Unfortunately, I never got involved in robotics in high school. I really wish I would have, though I was raised in a rural area where robotics was not as big.

      That being said, I had experience in embedded systems before taking this course, but it is really tailored towards any experience level. This project has a lot to offer for both people that are completely new to robotics and those that have been veterans for years. My avionics team had little, if any, robotics experience going into this, and we came out quite successful. Experience in FIRST Robotics would definitely help you out in a project like this; I know there are quite a few people on the team that have participated in it. It would definitely reduce any learning curve and allow you to contribute more.

      If you are interested in Carnegie Mellon, I highly recommend taking summer courses in EE or CS or taking part in a high school robotics program there. If you like robotics, CMU is the school for you! :-)

      Delete
    2. Thank you for your helpful advice! I'm sorry if I'm being slow but what do EE and CS stand for (I'm not in college yet so I don't know any fancy acronyms yet)? Our robotics team has three subteams: the electrical team, the mechanical team, and the programming team. I do not yet have enough information to figure out which one of the three I like the most. Do you happen to know anything about the differences between these three teams as an experienced college sophomore? Or are they all the same?

      Also, I notice from your profile that you are a Bay Area intern and in your profile picture, you are standing in front of 1 Infinite Loop. Does this mean you are an intern at Apple (the one that makes Macbooks and iPhones and such)? If so, that's so cool! What's it like and what do you do? How much would you say that working on the Tetramorph has prepared you for life as an intern (presumably an engineering intern)?

      I'm sorry if I'm asking so many questions. I'm just so interested in robotics and STEM!

      Delete

 

Recent Posts

Audience

Follow by Email

Copyright © Brandon’s Blurb | Designed by Templateism.com