During my first term in Mechatronics Engineering at the University of Waterloo, I've been pretty busy with robotics! One of the things I decided to do with what little free time I had was to compete in the University of Waterloo Robotics Team (UWRT) mini-sumo robotics competition during November of 2010. The team consisted of some familiar people which were a part of my "Jack the Blackjack Playing Robot" team. This time, our team's name was "Honbasho", named after a Japanese sumo-wrestling competition. We managed to place first in the competition against 13 other teams, This page is about the robot and our development/competition experiences:
Our prize winning 500g mini-sumo robot
The purpose of the mini-sumo class robot is to push the opponent out of a circular ring with a detectable border. The robots are constructed from predefined components, but the chassis design and the placement of provided components is up to the team. With the robot detailed here, myself and 2 other team-members who are also attending the University of Waterloo placed first in the UWRT sumo robotics competition. Some of the major components of the design are described below.
Here are some more pictures showing what the design looks like.
(and some imperfections I would've liked to fix before the competition!)
The robot's chassis is based upon a single piece of aluminium sheet metal which all teams were distributed. Because I had experience with this type of material and with AutoCAD design software (see the harmonograph table
), I was tasked with designing the robot's chasis so that it could be folded out of this sheet, as if from paper. After some sketches, I made the following drawing to aid construction: (click for larger view)
The plans for the robot were very detailed to avoid hard-to-fix mistakes in the sheet metal.
Each team was provided with a very capable controller board. The board is based upon the popular Atmega168 IC which is interfaced with the also popular arduino IDE. The boards came as bare PCB's and portion of the construction time was spent getting them populated in the electronics lab. After a couple hours of intensive SMD soldering, we had a working micro controller board for our robot! The board runs with a dedicated PWM chip for the motors which kept itself cool via a board-based heatsink. The board schematics are viewable here: (link)
We were provided with the following sensors to use in the design:
|Part name.|| Part description:|| How used?|
|1 x Sharp GP2D120 IR range sensor||This sensor uses IR light and optics to determine the distance between the sensor and another object in the path of the 'beam'||We used one of these on the front of the robot (attack with the front plow) and another on the front corner of the robot to determine where the other robot is precisely. (see diagram)|
|2 x QRD1114 IR transistor/LED package ||These sensors contain an IR-LED and IR-Transistor package which are designed for much more up-close work. but are based upon the same principals as the larger range sensors||We mounted these over holes in the floor of the chasis looking at the ground, they were able to determine when our robot was on the edge of the sumo robotics ring (white's resistance is different from that of black)|
We were also given some small easily-mounted geared motors, battery packs, wire and assorted components which we needed to utilize the controller board.
Drawings of the motor produced with use of vernier calipers.
We wanted to produce a design that would take advantage of common sumo-robot tactics as well as make the best use of the provided components. Because each team was provided with two ranged IR sensors, it was decided that mounting one sensor on the front of the robot, and the other on the left side would allow our robot to look towards the center of the circle as it navigates the edge of the circle.
Here are some diagrams of what we came up with.
Constraints And Criteria:
called for the following tough-to-meet criteria:
- smaller than 15cm x 15cm at max. horizontal section
- must not weigh less than 500g
- must not be kit built (chassis must be own design)
- completely autonomous
- battery powered
- no dangerous material
- magnets may not be used (darn, no accelerators!)
- Good stability
- High traction
- no parts sticking out
- have a large horizontal component in pushing force
- Wide wheelbase
- Largest possible mass
- Wheels covered by an outer shell
- Bent pushing blade to provide lift initially (to destabilize opponent)
and then horizontal pushing force.
After a couple weeks of development and some late-nights at the student workshop, we finished our robot! Of the 20 registered teams, only 13 made it to the competition (a new record we're told). After some good competition, we managed to remain undefeated in all our matches and went on to claim first place! I produced a video documenting the competition:
The results, we were undefeated!