2.2mb AVI

Question 1: What was the general strategy or plan behind your bot design?

The idea behind Mini-Skirt was to "trick" the enemy bots that use a light sensor to detect the edge into thinking they have seen the edge when in fact they have seen the "skirt" that borders 3 sides of Mini-Skirt.

As far as a strategy goes, there really wasn't one. I thought of the skirt as a passive defense to protect my most vulnerable sides, so I didn't program Mini-Skirt to use it in an aggressive manner. In addition to this, I had built a front bumper that had counter-rotating wheels and geared the drive train down to 24:1 for strength in a pushing situation.

Question 2: What technologies/mechanisms did you use?

The front bumper was directly driven by a motor which was directly connected to a 9v battery box. The output axle on the motor was extended to reach the front end of the robot to drive a worm gear while allowing the motor to be mounted in the rear where the rest of the weight was. The housing for this worm gear had to be constructed from scratch in order to be able to mesh with a 16 tooth gear. The gearing on the bumper was designed to be slightly faster then the gearing that the drive train used (a suggestion from Mario Ferrari).

Another mechanism employed was the powered wench that controlled the deployment of the skirt. This is basically a string spindle element driven by a motor that is controlled by the RCX. The strings used were braided nylon (almost exact copies of the standard LEGO string except in color) that I found at Home Depot. I used string instead of a more mechanical means to deploy the skirt as a way to save on weight.

Question 3: What was the most challenging aspect of construction?

I built and rebuilt the whole skirt mechanism 3 times before settling on a solid design. The challenge was in the pivoting of the hinges that held the skirt. Those on the rear of the robot are perpendicular to those on the sides of the robot. After a few trial-and-errors I was able to route my string in such a way that a single motor and spindle was able to reel in all four lines simultaneously to allow for a smooth deployment of the skirt.

Question 4: Are there any special features you care to talk about?

Although it didn't actually help me win any matches, I was especially pleased with how my physical edge detection sensors ended up being built. Again, some trial-and-error allowed me to see that my first design wasn't getting a solid contact on the ring surface. So I had to rebuild the sensors to get a better touch. What ended up being a bad thing was that any sort of vibration of the robot or the ring caused these sensors to trip. Of course if I had a better program, I might have been able to ignore false readings.

An additional component to the edge detectors that I really liked was the fact that my light sensor was originally mounted on a lift arm that was attached to the robot via a technic pin. This allowed the lift arm to pivot freely. This was very clever because it allowed the light sensor to stay the same distance away from the surface of the ring even if the robot was being lifted up until the lift arm has swung it's full stroke. The idea was that I would be able to check the light sensor in a situation where both edge detectors reported being off the edge. If the light sensor reported a value equal to or less then the initial value of the black surface then the robot was being lifted up and should take evasive action. If the light sensor reported anything higher then the initial value then the robot was on the edge. Sadly this whole assembly was removed prior to the event due to being over weight.

Question 5: What surprised you about how your bot behaved/performed during the event?

I was skeptical of the effectiveness of the skirt when I built it. As it turned out, most if not all of the opponents that used light sensors to detect the edge were tricked by it. In fact, it was so effective that I believe Mini-Skirt may have actually beat or at least tied with Knife if Steve hadn't disconnected it's light sensors prior to our match.

Also, during weigh-in I was surprised that I was about 3 ounces over weight. It took a bit of disassembly to come down to an allowable weight.