Whether it be social, recreational, or professional, some of what represents me is here. Post a comment, or contact me at should you so desire.

The posts are in reverse chronological order, and are pegged by topic on the links to the left. For more of an introduction, please see the About this site page listed above.

Sunday, 12 January 2014

Robotics Competition at the ISU

The first week back at the ISU has been extremely busy and extremely fun! Thanks to our special guest lecturer Professor Kazuya Yoshida, one of Japan's leading robotics experts, as well as the cooperation between the ISU and ENSIIE (École nationale supéerieure d'informatique pour l'industrie et l'entreprise), we had a week of robotics concluding with a robotics competition!

First, we had a few lectures on the prevalence and importance of robotics in space. A brief history of the concept and development of robotics was discussed, reminding me of how much science fiction I have read in my life. An interesting factoid for you readers out there concerns the origin of the word robot. A Czech word meaning worker, or servant, it was originally used in a play by Karel Čapek in 1920. Since then, there have been numerous stories of automated workers and human-like machines, most notably the works of Isaac Asimov, whose work has been made into the popular films Bicentennial Man, and I, Robot.

In reality, robots have been developed and used most notably in situations which prove too difficult or dangerous for human beings. In the space industry, robots have been used to transport and move equipment, sense the environment, and move on other celestial bodies. All of the rovers and landers used on the Moon, Mars, and other planets are considered robots and have a good history of helping us learn more about our solar system. Some examples of these include the Viking Lander (landed on Mars in 1976), the Mars Curiosity rover (landed in 2012), and most recently, China's Yutu rover on the Moon!

As a Canadian, I am especially proud of the history and use of robots in the space industry. The Canadarm, first installed on the Space Shuttle Columbia, proved so useful, it became a staple of the Shuttle program. The Canadarm2, installed on the International Space Station, was exceptionally useful for moving and installing the various sections of the station. If you're interested in trying your skills as a Canadarm operator, try out this game, provided by the Canadian Space Agency!

After the lectures, the students at the ISU were given the chance to team up with our sister school, ENSIIE, working together toward designing and building robots! The objective? Build a robot capable of moving throughout an environment, avoiding obstacles, not going over the boundaries, and collecting as many gems, or marbles in this case, as possible. Meant to not only be fun, the exercise was to help illustrate the challenges involved with creating robots which could perform a task on their own. Many rovers on other planets perform functions autonomously, on their own, such as moving around and avoiding obstacles, so this was an exciting opportunity!

Separating into our separate design rooms, we were given a chance to meet each other. My team comprised of 4 French students, two of my colleagues from the ISU, and myself. Everyone on my team understood English, although the French students and one of my colleagues would often lapse into French. While I haven't had much of a chance to get out and explore France these past few months, I will say that I am proud that my ability to understand French is improving. I find that, as long as I'm paying attention, I can almost always understand exactly what is being said. Responding is a little slow and error-prone, but I am working on it!

One of my colleagues had taken the challenge back in the ISU's Space Summer Program so he said he would sit back when it came to design, not wanting to influence our end result too much. Opening our robotics kit, it seemed we would be using the LEGO Mindstorms kit, which makes me jealous of the kids of today. When I was growing up, the most complicated LEGO set we had contained merely plastic pieces and maybe a few decorative ones, but nothing like this! Closer resembling the K'Nex set I had as a child, we had two full boxes of plastic connecting pieces, as well as an ultrasonic sensor, two touch sensors, a light sensor, and a programmable computer!

Excited to get started, many members of my team cataloged the pieces, making sure we weren't missing anything.

While this was happening, I began working on the design of the robot. As I saw it, there were several subsystems, parts, of the robot which had to be considered in order to ensure we built a successful robot.

These subsystems were:

  • The Sweeper - the part of the robot which would gather the gems
  • The Storage bay - as we were required to not only gather, but collect the gems off the floor
  • The Driving system - we had several choices of wheels as well as a tank-like track system
  • ultrasonic sensor - would be used for obstacle avoidance
  • 2 touch sensors - which would also be used for obstacle avoidance
  • 3 actuators - motors which could be used for driving the robot and the sweeper system
  • light sensor - which would be used to detect the white line of the boundary and keep the robot inside, also could be used to help the robot stop on the black starting mat, a condition which would stop the challenge but double the points per gem collected
Some of our early designs
Beginning with the Sweeper system, I came up with a few designs, and then consulted with the team. As you can see, we had several interesting ideas. The first was simply a sweeping, rotating wheel which would would move the gems into the storage bay. Another idea was some form of catapult to launch the gems into the bay. Next, there was the idea of the conveyor, using the treads to move the gems up and into the bay. Finally, we had something known as the "reverse sweep" which would use two claws or wheels to pull the gems into the bay.

We were lucky as we were also provided with the LEGO Mindstorms book which provided reference for making a driving system, and a wheel which could allow the robot to turn in any direction, similar to the wheel on a shopping cart, only better. Following these guides, we created a simple driving system, just to get an idea of the size and pieces we required.

We then constructed a storage bay, and attempted to attach it to the front of the driving system. However, this proved too brittle of a design, so we took the driving system apart and decided for a more vertical approach.

Choosing wheels over treads for their simplicity, we separated them on either side of the storage bay and added a roof.
Our little LEGO man decided to help us out!
Following that, we moved onto the Sweeper system. At a funny suggestion from one of our teammates, we decided to use three claw-like pieces found in the kit and attach them to an actuator. The actuator would cause the claws to spin around and around scooping up any gems in its path. This sweeper would be added to the top, front of the roof, with a wedge-like design underneath to guide the gems toward the sweeper as the robot moved forward.

As mentioned in previous posts, and as anyone who has ever designed anything will know, each choice made affects future choices. The sweeper in the front meant that the ultrasonic sensor had to be moved slightly to the side. Normally, the ultrasonic sensor would be placed in the centre-front for maximum effectiveness at detecting the obstacles, but with the claws spinning around and around, the sensor had to be moved to avoid detecting the sweeper.

Everyone agreed that the touch sensors would be placed on the side, facing the front, and would react to the robot hitting an obstacle out of range of the ultrasonic sensor. We added some objects to the touch sensors so as to extend their reach. The light sensor would be placed to one side near the front.

While all this structural engineering was taking place, one of our team members was working on the computer program. This program had to respond to the ultrasonic sensor and touch sensors to avoid obstacles, reverse and turn to avoid the white lines of the boundaries, as well as driving the wheels and sweeper of the robot. I did work too much on the program's design, but it was a very integral part with a high degree of risk; as one wrong command could cause the robot to escape the white boundary and be disqualified, and the logic of the programming had to be worked out in order to promote success.

Putting all of this together, and running several tests, we finally came up with...the WALLverine! (A combination of WALL-E and Wolverine, because of the claws)

This was the design we had been using for some time, until the arm became slightly unstable and started to droop. Here, you can see the ultrasonic sensor (looks like WALL-E's eyes), sweeper claws, the two wheels in the front are attached to the touch sensor.

Noticing that the sweeper was starting to droop as it was working, we had to do some work strengthening the support. This support caused the claw sweeper system to become stuck, so we ended up removing one of the claws. (Which made me feel better anyway because Wolverine has three claws, not four, after all) Testing it out, the WALLverine was ready to go!

As were the other robots. They were placed on the table, and with everyone crowding around, I attempted to get some pictures, although I did miss taking photos of some of them. An interesting aspect of these types of challenges is to note the similarities and differences between the robot designs. Faced with similar pieces, with similar design constraints, some of the robots were similar. However, since most teams were fairly isolated, the designs were different enough as to be interesting.

I really enjoy the claws and, while hard to see, the tread conveyor system.
Similar to our WALLverine, this used small plastic pieces for its sweeping system.

I felt this one looked like a combine, a grain processor, and that this view is pretty intimidating. 

This robot turned very quickly, and used a rapidly spinning wheel system for its collector. 
Another interesting design, combining the conveyor system, with a different take on the structural support.
A simple looking, but robust design, this robot featured Admiral Ackbar from Star Wars. Additionally, anytime it sensed an obstacle, it played "It's a trap!" through some speakers.
 A view of the obstacle course. The white lines were white duct tape, the obstacles were concrete blocks, and the gems can be seen, white, orange, and scattered all over the floor.
A bird's eye view of the course. The black square is the starter area. Bonus points for returning there with gems.
Fit with sound-system and microphone, the competition started off and was commentated on by Prof. Yoshida. Points were given to avoiding obstacles without touching them, and points were given for collecting gems.

In the space industry, rovers like Mars Curiosity roam over the Red Planet's surface looking for things of interest, namely rocks and soil. While the current robots, like Curiosity, collect and analyze the material in-situ, in place. some interesting developments in these types of missions. Known as a "sample return", the robot collects the material and brings it back to Earth for analysis. While a more complicated and risky mission, the main advantage is that scientists here on Earth can analyze the material first-hand, which usually allows for better analysis.

The reason I mention this here is that we were given two trials, taking place in 5 minutes time. During this time, the robot would move on its own collecting gems but, as mentioned before, bonus points were given for bringing the gems back to the mat, mirroring the importance of sample return missions.

The competition was pretty intense, and all the robots were really exciting to watch! Some were slow and thorough, some were fast and retraced their path. All of the robots did really well, and the crowd was cheering.

Here is a short video of our robot in action, showing it avoiding a boundary and an obstacle. I imagine it was really funny watching me while the robot was moving.

Each team selected a "pilot" who would place the robot at the starting mat and be responsible for stopping the robot should it become stuck. While the other teams's pilots would place the robot on the mat and stand back and watch it go, I followed the robot around the track, cheering it on, and even pointing in the direction I wanted it to go. Because I had done this during the testing, I had a fairly good idea of where the robot would go when it avoided obstacles. To everyone else, I must have looked like I was directing the robot.

We ended up only using one of our time trials and returning to the mat. Our only main problem was that the sweeper was turning too quickly and was accidentally throwing some gems out as it turned. We did not have a chance to test this beforehand because 20 gems was the maximum we had been allowed to test with, compared to the 100 or so which were placed on the obstacle course. However, we were very pleased with the results as we placed 3rd!

While all the robots did well, two robots in particular greatly outperformed the rest. The Admiral Ackbar robot moved well, and its collection system worked flawlessly, since it returned to the mat, it ended up earning over 500 more points than our 3rd place.

The winner of the competition was the one featured above with the "rapidly spinning" collector system. It moved quickly, turned very smoothly in place, and just barely scored more points than the Ackbar team. I have to say that this robot was the most exciting to watch because it circled around and around the starting mat, collecting more gems, and as the time was counting down, the anticipation rose as we wondered whether it would make it back to the mat. Collecting over 60 gems, and just touching the mat as the clock struck zero, the entire room erupted in applause as Harvey the Harvestor (as the team was calling it) touched down on the mat!
The final results!
The winning team, featuring Professor Kazuya Yoshida (in the middle), ISU President Walter Peeters (wearing black on the centre-right), and Professor Hugh Hill (on the far right). 
The winning team was presented with gift bags full of ISU and ENSIIE memorabilia, as seen above. Afterward, the teams were invited to put their robots back in the field in an ultimate fighting showdown!

At first, our robot was too good at avoiding everyone, owing to its touching and ultrasonic sensors. Deciding to try something new, we removed these sensors and sent our robot back into action! Turns out the claw system was a little too effective as our robot would latch onto other robots and not let go.

While trying to pull our robot off of another, the WALLverine came to a tragic end, ripping itself apart.

And that concludes our robotics competition! It was an exciting couple of days, and the students from the ISU and ENSIIE agree that it was a fun collaboration and hope for more in the future! I hope you enjoyed following along in our adventures and come back in the future for more exciting developments from the International Space University!

Thanks for reading!

No comments:

Post a Comment