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 25

Handy Board troubles: we were having trouble testing out the magnets, Handylogo kept saying “handyboard not connected.” When we finally got the handyboard to respond, it was making weird clicking noises when the program was running. We restarted the two boards that we had to see if that would help. So far the handy boards seem fine.

We hot glued the magnets into the treasure chests. Each of the five chests got two large magnets, glued to the inside of the top of the box.
We did a demo for Professor Berg. The car can find the treasure fine, but he’s not backing up enough before he digs. He also sometimes senses the treasure from too far away – the treasure is off on his left or right but he beeps and starts digging anyway. We can fix the backing up thing easily – we just increased the time he went “thatway” with the motors on. The second problem required a little more programming. Since we have two sensors, a left and a right, we can kind of tell which side the treasure is on. So Stephanie modified the code so that if the left sensor found something, the car would turn a little to the left and then back up and dig, and vice versa for the right sensor.

The underside of our car - The white circles are the two magnet sensors.

Karen and I worked on the bumper. Professor Berg suggested that we attach the bumper differently so that it hinged on less of an angle, so we tried that. At the moment, the car doesn’t deal very well with walls. Most of the time, it doesn’t even set the switch off, b/c the bumper doesn’t connect to the switch very well. Hopefully the change in angles will help with this.

Car with modified bumper attached.

The car also had trouble digging up one of the treasures because of a slope in the sand – the brush was digging too far into the sand because there was kind of a small hill underneath it, between the wheels. Some of this can be solved by adding a little more sand so that the treasures don’t cause such a lump/hill in the sand. Professor Berg suggested that we add more sand, and then use the edge of a board or a broom to even out the sand after the treasures had been buried. We’ll try this next time.

Another way to deal with the problem of the broom dragging in the sand problem was to attached a touch sensor to the top of the “broom” so that we could tell how many times the broom had “swept.” Stephanie programmed in a global variable that kept track of how many times the brush had swept (how many times the sensor had been touched). We programmed it so that the brush would sweep 3 times and then hopefully stop somewhere near the top of it’s arc.

Once we came back from lunch, we spent some time working on a bunch of different things. We made parts of the car more secure, especially the tower on the front. We decided that the slope of the sand was giving him some trouble, so we filled the entire frame with two or three inches of sand. We made slight adjustments to the timing and sequencing of the code, so that he found and dug up the treasure in a more accurate way.
We also modified the behavior of the car once it hit a wall. Before, when it hit a wall it would back up for some random number of seconds between 12 and 16 or so. But the car wasn’t doing this for some reason – sometimes the time that it backed up for was barely a second before it started going forward again. So we just decided to have it back up for 10 seconds each time. This was enough time to get quite far away from the walls to start randomly searching again.

We discussed the look of our project. At that moment, he looked like a giant red ant, which didn’t really appeal to any of us. The layout of the car didn’t really lend itself well to a carriage or a pirate ship, which was what we had been planning. We decided that it kind of looked like a dog – the front bumper part being his “snout.” Since he was already mostly made up of red legos, we quickly came up with the theme of “Clifford the Big Red Dog goes to Treasure Island.” It was slightly different than our original plan, but it worked out ok. We rummaged through the craft supplies and found some red felt. We covered Clifford in the felt and added ears, a tail, eyes and a nose. This was good because it covered all the unsightly wires while not getting in the way of any of the mechanical parts.

Clifford in the process of being decorated.

We showed our creation to the class, but we think Clifford got a little stage fright, because he kept running into the wall. We realized later that we had plugged the front bumper sensor into the wrong port. We decided to prevent future mistakes along those lines by tagging each of the wires with the motor or sensor it belonged to and the sensor or motor port that it was supposed to be plugged into.

We also added painted the treasure chests a mix of brown and gold, and created a simple decoration of palm trees to cover the unsightly electrical tape and plastic on the outside of the sandbox frame. We plan to decorate the treasure chests, move the sandbox, and fix up any last minute bugs when we meet tomorrow morning before class.

A couple pics of us working on Clifford during the day:

Clifford, all decked out and ready to go: