Construction Commences!                                                                                                          Saturday July 16, 2011
Moonbots 2.0 Challenge Team 364

Time until Phase Two ends:

Today we had our first major meeting to start constructing everything necessary for the robot portion of MoonBots 2.0. We started to set up the field and figure out how to approach the challenge ahead of us. Overwhelmed and excited by the tons of LEGOs we received, we quickly went to work. We started with the landing ramp...

...and then onto the ridges of the craters. Joey decided to take a little break:

After we assembled all of the ridges, the helium, and the water ice, we went on to look for the pieces for the Peak of Eternal Light, which the robot must visit in order to “survive the lunar night.” Unfortunately, we were missing some of the pieces, but we hope to get them soon. We then started setting up the field, piece by piece.

We then started toying around with a quickly made practice base (of no particular moon-themed design) to get an idea of the scale of the robot and figure out what the optimal path for the robot would be to traverse the field. We’re aiming to complete all of the mission objectives while encountering the least amount of lunar obstacles.

After our long work day, we sat down with some iced tea and discussed the details of our upcoming STEM outreach project. We have some great things planned for this Wednesday, including an interactive space-related presentation and a demo of our FRC Robot.

View full-size images and more pictures in our Photo Gallery.

Quick Update!                                                                                                                               Tuesday July 26, 2011

    We took a short break while we were waiting for our missing pieces to arrive (we just received them today!) and while some of our group members were competing at the Indiana Robotics Invitational this weekend. In related news, the two week long Raider Robotix summer LEGO camp concluded this past Friday! We’re still in the process of compiling the pictures and video we took from our presentation. Finally, we’re planning to meet up at Tommy’s house this weekend to continue working on our robot!
STEM Project Overview                                                                                                                Wednesday July 6, 2011
MoonBots Game Day!                                                                                                                 Wednesday July 20, 2011


Today the LEGO campers got to learn all about MoonBots! Sue, Tommy, and Joey gave a presentation, starting off with a game of Hangman where the puzzle was “Google Lunar X Prize!” The kids were excited to know what this was and so we talked to them about what the MoonBots and the Google Lunar X Prize competitions entail and why they are important. The kids were so excited to learn space exploration! After the “lecture” portion of the presentation, we put the skills of the campers to test in a fun game of Jeopardy!  In the end, the campers learned a whole lot about STEM education, NASA, engineering, and the Google Lunar X Prize. The The winners (everyone’s a winner when it comes to learning) got some nifty MoonBots stickers and a sweet treat! Watch out for the full coverage of our STEM presentation coming soon!

UPDATE: Our STEM Page is now live! Click here to check it out! There’s more detail, pictures, and video!

Another Build Day! [New Pictures and Video!]                                                                             Sunday July 31, 2011

        In order to rapidly test the mechanics of our prototype without having to spend time programming, we simply wired two NXT motors together and attached a crank handle to one. When the handle was spun on the first motor, the second motor would also move. Even though the movement provided by this method was slower and shakier, the pieces still made it into the robot's hopper, so we're confident that the faster and smoother motion obtained via programming will be just as, if not more, successful.

We've received the missing pieces that were required to build the Peak of Eternal Light, so building that was the first task on today's to-do list.

        Then we began prototyping and testing the design for our robot's end effector. The "element scooper," as it was dubbed in our robot design proposal, needs to be able to accurately and quickly pick up the Water Ice and Helium 3 element game pieces so that the robot can bring them back to the landing base for bonus points. Our initial design proposal was based on the design of the game pieces from last year's Moonbots competition. However, this year the challenge is harder because the elements are not simple rings that can be speared through the center. To accommodate for this change, we decided it would be best to pick up the pieces from below. The elements' narrow bases would fit in between the prongs of our element scooper and catch on their wider tops as the claw is lifted.

        As Sue and Joey set up the Moonbots playing according to the official blueprints, Tommy tweaked the spacing between the rakes on the element scooper to account for the distance between each of the elements. Our game plan is to pick up the Helium 3 elements in pairs first, travel to the crater Chanda to pick up the pair of water ice, and then go to the larger crater Canvin and collect each of the water ice individually since those are spaced further apart and our robot will have to travel over the regolith. Because the water ice in the crater Chanda have a greater distance between them than the Helium 3 elements, and the robot has a starting size restriction of the size of a sheet of paper, we are planning to add extra rakes on the sides of the element scooper that will fold down and out at at the start of the mission. We employed a similar technique in the design of our 2007 and 2011 FRC Robots to expand beyond the imposed starting configuration, so the lessons learned from those experiences will prove to be valuable here.

        We then constructed a hopper for the robot to deposit the game pieces in after it has scooped them up. The hopper will be located on the topmost part of the robot and bot's claw will do a sort of "reverse slam dunk" to deposit the elements for safe keeping.  We had to make sure that the hopper was large enough to fit all ten elements without falling out. The walls constructing the sides and back of the hopper are slightly angled outward, like a funnel, so that there is a greater area in which to catch the game pieces. We did a lot of tweaking to the front wall of the hopper, as sometimes the elements would fall off of the claw a bit too early. We remedied this problem by making the front wall lower and moving it closer to the front of the robot. Furthermore, sometimes the game pieces would get caught on the front corners and just barely miss making it into the hopper. We solved this by rotating the LEGO pieces that were in the way about 90 degrees around the axles they were mounted on so that they were no longer in the way but would still guide the pieces into the hopper.

Check out the photo gallery for even more pictures from today’s session.

Older entriesBlog_%28Page_1%29.htmlshapeimage_17_link_0
Older entriesBlog_%28Page_1%29.htmlshapeimage_19_link_0

    We’re proud to announce that our STEM outreach project will be Raider Robotix LEGO Camp, a two week day camp that will run from July 11th to 22nd. We currently have seventeen kids registered! The camp will introduce the campers (ages 10-14) to fundamental engineering and programming concepts using the LEGO Mindstorms platform. We have a lot of fun things planned for those two weeks, including a special MoonBots Day and a demo of our 2011 FRC robot. We know the campers will get a lot of hands on experience as they work their way through the tasks we give them each day, whether it be gearing a robot to win a race, or programming light sensors to complete an obstacle course. Here is the preliminary basic lesson plan (more detailed one to follow):

Monday (July 11)
Briefly introduce the Camp Program Team; assess this year's crop of participants and break them into teams; get them started building bots for initial races, coaching on the importance of strong builds that will sustain the rigors of competition; robot drag races throughout the day.

Tuesday (July 12)
Introduce the math behind gearing to illustrate the tradeoff between speed & power. Robot racing continues as campers become more comfortable using LEGO Mindstorms. Introduce other simple challenges such as using geared power to push things, climb, or overcome other obstacles.

Wednesday (July 13)
Robot Sumo! Low gearing is your friend here! The robot will the most powerful drivetrain and traction will come out on top.

Thursday (July 14)
Sumo continues in the first half of the day; possible robot demo in the second half of the day. (More on that to come.)

Friday (July 15)
Introduction to programming and sensors.

Monday (July 18)
Sensor and programming continues.

Tuesday (July 19)
Navigating mazes using sensory inputs.

Wednesday (July 20)
Line following using light sensors.

Thursday (July 21)
The Grand Challenge—to be designed.

Friday (July 22)
The Grand Challenge continues and concludes in the first half of the day; participant surveys; pizza party in the second half of the day.

Each day will bring a new challenge to solve. The camp will be taught by a group formed from members of the Raider Robotix MoonBots and FRC teams. This is a huge project for Raider Robotix to pull off, but if we put our minds to it, nothing is beyond our reach!

UPDATE: You can now view the complete schedule on our STEM page for LEGO Camp!

Sensor Testing And Build Progress                                                                                                     Sunday August 7, 2011

In our robot design proposal, we suggested the use of two different sensors: the magnetic compass sensor and an ultrasonic sensor. We decided to collected data from each of these devices to test their accuracy so that our robot would not be thrown off course due to faulty readings. Here are our results and analysis:

First the magnetic sensor, which would output a heading from 0 to 359 degrees based on the earth’s magnetic field, was tested. The sensor, as far as could be told, displayed values within the proper ranges. For example, when the sensor was pointed towards the true north it would read 0 (as verified by an iPhone compass). When it was then rotated 90 degrees, the reading was somewhere in either the high 80s or 90s, even slightly surpassing the 100 mark. The sensor was very “jumpy” and would change with the slightest movement of the robot. Additionally, when another electronic device (including one of the LEGO motors) was brought near to the sensor, the readings would continually change to wrong values. It was deemed that this sensor had too low of a tolerance to be effectively used in our competition robot design. A program could have been developed that would allow the robot to detect fluctuations in the sensor’s reading, but we deemed this to be too complicated to carry out in the standard LEGO NXT programming language given our time constraints.

Next we collected readings from the ultrasonic sensor. First, a simple program was developed to display the sensor’s current value to the 100x64 pixel screen of the NXT controller (see the picture to the right). Then, using a meter-stick, we recorded the sensor’s readings at 1 cm and 1 inch intervals, and calculated the the error between the actual and sensed values calculated. 

We then plotted the data from the table (top left) to obtain the above graph. We noticed that when using inches to report distance, the ultrasonic sensor had a lower error because the scale of English units is greater than metric. The average error, when measuring in centimeters, was about 2 cm. The sensor was generally better at reading farther distances compared to nearer ones because the error reduces as the distances increase. This phenomenon can be seen in the negative slope of the graph. This could be because the waves emitted by the sensor are disrupted by collision interference at closer distances. We determined that the maximum distance threshold for this sensor is around 86 cm, while the minimum threshold is approximately 2 or 3 cm because the error in the close range was greatest at 1 and 2 centimeters. These tolerances meet our requirements, so we will be using two ultrasonic sensors on our robot, facing ninety apart so that they don’t interfere with one another. This will allow us to get readings from the walls bordering the field to the front and right sides of our robot.

In other build related news, we’ve constructed the base and drive train for our robot. We originally proposed using a ski to support the front of the robot, but found that there was no feasible way to create such a design that could easily cross over the craters/bumps on the field. Therefore we have decided to place two smaller, unpowered wheels at the front of the robot to aid in maneuverability. However, we are still experimenting with the design and type of tread for these wheels. We’ll report back next week.

Work also continues in designing the robot’s CAD using the LEGO Digital Designer and editing our STEM presentation video.

The Final Countdown                                                                                                                     Sunday August 14, 2011

This weekend we finished building the robot’s base and started programming for our live mission broadcast that is scheduled for Tuesday, August 16th at noon EST.

In the last blog post we mentioned that we were trying to decide what type of wheels should be used for the front of the robot in order to help our robot get over the crater ridges while still being maneuverable. We did not want the front wheels to have any type of tread on them, as this would increase traction and scrubbing while turning, as well as effectively increasing the length of our wheelbase. Therefore, we aimed to use some sort of slick wheel that would act more like a caster since it would be unpowered. First we tried using some pretty standard white wheel hubs. However, we quickly realized that the wheels’ outer ridges (where the tread would normally attach), would frequently get caught on the pegs of the surface of the placing field, hindering the robot and messing up its turning. After searching through many of Tommy’s LEGO containers, we struck gold when we found small, black wheels. After swapping them out, the robot moved like it was gliding across water ice. ;)

You may be wondering what the treads on the underside of the robot are for. These treads are powered, but have enough clearance that they do not touch the playing fields surface or the moon rocks. However, when our robot climbs the crater ridges, these treads propel the robot up and over! Thus, our robot is able to maintain a low center of gravity while still being able to surmount any obstacle in its path.

After construction was done, we began modeling the robot using the LEGO Digital Designer and programming portions of our code. We started by making functional blocks, such as turn 90 degrees left/right, and arm up/down, and then chained them together using feedback from the rotation sensors in the motors and our ultrasonic sensors.

You’re also probably wondering where our STEM presentation details are—don’t worry! It is being worked on behind the scenes and the dedicated STEM web page will go live soon! The video and other elements are in the final editing stages.

As always, more pictures from today’s meeting (and higher-res versions of the ones in this post) can be found in our photo gallery. Finally, here is a very short “teaser” video of our robot. We’ll update again after Tuesday’s mission broadcast!

Mission Complete!                                                                                                                          Tuesday August 16, 2011

To start, we’re proud to announce the launch our STEM outreach hub! Learn more about LEGO camp, MoonBots game day, and the FRC demos we did to share our excitement about robotics and technology!

Second, today was our live mission broadcast. We met James, the MoonBot admin, via Skype to talk to him about our STEM activities, robot design, and have our robot’s routine be judged.

On our best trial, we scored 140 points by getting off the lunar dismount, collecting two Helium 3, photographing the heritage artifacts, and returning the base, all while taking video with our onboard camera. Here’s the footage that was captured:

We wish that we would have had some more time to work out all of the kinks in our programming so that we could have accomplished some more of the tasks that our robot was designed for and capable of doing, such as climbing the crater ridges and navigating to the peak of eternal light (as can be seen in this video).

Overall, participating in the MoonBots 2.0 Challenge was a positive experience for our team, and kept us busy in between all of our summer activities. We look forward to finding out who the winning teams are for this year and competing in next year’s challenge!

Final Robot Design                                                                                                                          Monday August 15, 2011

Drive Train:

Although simple, our robot’s drive train is actually very elegant. The simple direct drive system powers the two wide back wheels (Which are not pictured in the final CAD because LDD did not have them in the available parts) as well as the two treads that

run under the center of the robot’s mass. While the treads do not touch the ground on level ground, the treads contact the surface of the ridges when our robot attempts to cross the moon ridges and gives our rover that little bit of extra traction needed to propel it over the ridge. Moreover, as opposed to having two powered front wheels, our robot uses two simple caster wheels to keep the front of the robot supported. Their wide but slick outer rim ensures that they will slide easily across the fields surface and serve our robot better than the skids we previously tried (In the Lego CAD, not only were the casters not present in the list of parts, but the limitations of the program did not enable us to connect the caster mounts to the entire frame of the robot). Overall, what our rover’s drive train lacks in gears and complexity it makes up for in structural integrity and consistency.

End Effector:

Similar to our drivetrain, our end effector, which we have dubbed the “Element Scooper,” is simple but functional. This simple system needs only two parts to effectively harvest the elements we come across on the moon’s surface. As it drives around the moon setting, the simple fingers on our robot’s arm hook the elements under their rings and above their base. After this, when the arm lifts up to its highest position, the fingers pivot down and allow the elements to slide off them and into the hopper. Although the elements of this year’s competition are not as easily harvested as the simple rings of last year, our robot’s arm is able to capture them just as easily.

You can view more CAD models of our robot in the gallery! Some of the renderings even have fancy space backgrounds!