Clifford goes to Treasure Island

Design Journal

January
S	M	T	W	Th	F	S
16	17	18	19	20	21	22
23	24	25	26	27	28	29

January 24

Stephanie brought in the frame for our sand box, so we spent some time taping the plastic trap to the wood and pouring in the sand. We think only two of the bags of sand is more than enough (100 lbs of sand), especially since we don’t want the entire thing to be full – we want it to look a little more like an island – so the sand will be in a pile in the center of the frame, surrounded by “water” (blue paper mounted to the plastic underneath the sand).

We thought that thicker wheels would give us more traction in the sand, but we didn’t have any wheels that were thicker and had the diameter we wanted, so we doubled up – each “wheel” is actually two of the large wheels placed on an axel side by side.

We tried out our vehicle on the sand with the double wheels and a basic version of our “search” program that Karen had written. Basically, our program had the robot going straight until it hit a wall, then backing up and turning right for a little while, then continuing on straight again. At this point we just had a touch sensor mounted to the front of the car, so it wasn’t very sturdy. There were some problems with the code for this, and the sensor kept popping off, so we decided to work on the programming for a bit. We wrote a different version of “search” that turned and went straight for randomly chosen intervals of time – we figured this would be more fun than having the robot search the area as if it were a grid (like a lawnmower). We also realize that the robot would not really be able to search in a grid very accurately in the real world.

We hand some problems getting the robot to turn around when it hit a wall. It was having difficulty backing up the slope of sand and turning at the same time. Karen had the idea that we didn’t necessarily need to have the robot turn – it could just back up for a while and then continue it’s random search pattern.

Despite all our efforts, the robot was still having trouble on the sand. One problem was the length of the bristles on the brush we had attached were too long – when the car was driving across the sand it was scraping the bottom. We cut the bristles a little bit, but not very much since we didn’t want to cut off too much and then have to go buy another brush. We also decided to gear down a little more, the idea of which pained all of us. It wasn’t that bad in the end – gearing down the front motors simply meant replacing two of the 16 gears which had just been idling with an 8 to 24 gear down. The back wheels required a little more creativity. We moved the wheels back a little (away from the motor) and added in another 8 to 24 gear down.

We also moved the brush up a little bit, which also helped the car move better b/c it wasn’t dragging in the sand as much.

Karen worked on the front bumper – making it sturdier and more like the bumpers on the sci-borgs. We’re still having problems with it though – it has to be built around the paintbrush and also give the brush enough room to move and dig. It’s too high up, so it doesn’t connect with the walls very well.

We did some more debugging in our code (like telling the motors to turn off, then telling them to go thatway and wondering why the car wasn’t reversing). The searching works relatively well, but we think that there are still some bugs in it. We think the control dot is getting stuck in a loop somewhere b/c sometimes the car doesn’t respond to it’s front bumper (sensor) being pushed. We plan to talk to Lyn about it tomorrow.