One of the most fabled classes at MIT is 2.007 (previously known as 2.70) Design & Manufacturing I. In this class the students conceive, design, and create a robot which then competes in a final contest at the end of the semester. The class uses the contest as an excuse for learning all about design and manufacturing.
When I took the class the contest was named Ballcano. The basic point of the contest was to collect balls and deposit them in holes on your side of the table. The balls were being spat out onto the table by an active conveyor system at the rate of roughly 5 balls a second. For more detailed information about the contest you can look at the rules here.
The first thing that I did during the class was think about strategy. I tried to outline every possible robot that could conceivably be built, and I then tried to come up with strategies that could defeat each of those robots. Although there were several people who built robots that I never would have even considered, in general I was able to come up with most of the viable strategies.
There were several realizations that I came to early on:
- A robot which could capture balls from the top of the hill would have a great advantage over robots which were collecting balls at the base of the hill.
- A robot which could deliver balls directly to the holes would score much higher than a robot which had to ferry balls back and forth.
- Speed would be essential in beating other robots to the top of the hill.
- Almost all of the other robots would require precise positioning in order to function correctly.
With these functional requirements in mind I began to doodle out possible design ideas.
Based on the realizations that I had come to earlier, I began to develop certain key strategies.
- Get to the balls at the top of the hill as quickly as possible (along the top of the wall ideally.)
- Drag behind me some sort of ramp/conveyance system to transport the balls back to the 1 pt. hole.
- Try to divert as many balls as possible from the top of the hill towards my side.
- Build a second module which would drive on the ground and mess up my opponent.
These ideas then began to take the shape of a Wall Crawler, a Diverting Arm, a Ball Sock, and Wedge (The Annoyance Bot).
Throughout the design process I had to pay extra special attention to the physics involved with my robot, particularly the Wall Crawler. The torquing moment around the Arm threatened to tear my joint apart, and the torquing moment which the weight of the robot exerted against the wall threatened to pull the Wall Crawler away from the wall and out of position.
After resolving many of the physics issues I was able to begin sketching the designs of the components using actual kit contents, (although I was forced to go back and forth between the design stage and the physics as my robot evolved.)
I made a colelction of isometric sketches for each of the modules that I wanted to build. In addition, I also built numerous cardboard models which helped me to visualize the correct positioning of the modules, (the Wall Crawler and the Arm in particular.)
After determining the correct orientation and positioning for each of the modules I began to create finalized blue prints for each of the parts that I was going to build.
In the pictures below you can see each of the modules that I built for the robot.
- Wall Crawler Module
The Wall Crawler attached itself to the wall using a pair of pneumatic pistons. It rolled along the wall using the power of a Bosch motor attached to one of the kit wheels with a bura rubber surface attached to it, and on the other side of the wall a pair of delrin cylinders rolled along.
- Arm Module
The Arm Module was actuated by a Bosch motor which could be switched to either side of the robot depending on which side of the table I was competing on. The Arm itself was made out of shaped sheet metal with a lexan base.
- Ball Sock
The Ball Sock was made out of stitched cheese cloth with welding rod ribs sewn into it. It attached to the opposite side of the Wall Crawler from the Arm. Balls would travel from the Arm, across the Wall Crawler, and into the Ball Sock.
Wedge (The Annoyance Bot) was actuated by a pair of very heavy, very powerful Ford motors. It's wedge-like shape allowed it to force itself under other robots and either tip them over or pin them up against a wall.
The robot performed exceptionally well during the contest. The Wall Crawler module was able to consistently score 15-25 points per round, and the Wedge module was able to succesfully disable each robot that I went up against. This combination of offensive and defensive strategies proved unstoppable.