Beginning steps

Work in groups of 2.

1. Claim a pre-built robot (there should already be enough for each pair) or assemble one yourself. Instructions can be found on the class website. These instructions are generic. In addition, make sure that your robot has two light sensors, and that they are attached to sensor inputs 1 and 3. The two motors should be attached to A and C.
2. Have an TA install Program0 on your robot, to make sure it has been assembled correctly. When you run Program0, the robot will drive forward briefly, and then turn to the right.
3. Have the TA install the two behaviors that you will be trying to understand.

You may use any tools you wish to determine what causes the robots to do what they do. We recommend a flashlight and a white piece of paper. You should not disassemble the robot or otherwise do violence to it.

Experimental structure

For both Program1 and Program2:

1. For each program, first simply "play" with the robot. Write down a brief characterization of what you see. If possible write a bit of pseudo-code that you think could generate what you see.
2. Design an experiment testing this hypothesis. I.e., figure out (at least one dimension of) "independent" varying features of the robot's environment that you think will effect its behavior, and (at least one) "dependent" variable you can measure that captures an important dimension of the behavior.
3. Perform several iterations of your experiment. Be sure to do enough experiments with each behavior that you can tell how the context influences behavior. Does it seem consistent with your initial characterization? If not, how would you now describe the rules governing behavior?

General questions

1. In what ways are Program1 and Program2 the same? What are their differences?
2. Behaviors may depend on the context in which you turn on the robot. What features of the robots' environments have consequence with respect to any of the exhibited behaviors? Hint: context is defined by input values on the sensors.

Assignment writeup

Write a short (HTML format) document with answers to the questions above, using the following outline:

1. Program1

   1. comments about play...
   2. Your expt design...
   3. Results of your experiment...

2. Program2

   1. comments about play...
   2. Your expt design...
   3. Results of your experiment...

3. General

   1. Compare/contrast Prog1 with Prog2...
   2. Environment features...

Italicized text corresponds to your answers. In general your answers to each question will be approx. a paragraph in length, though it may take more to describe your experiments.

NOTE: if you can't write HTML then you may turn in a plain text file instead, with an extension .txt. No Word documents please.

Assignment submission details

1. Name your writeup behavior_<UCID1>_<UCID2>.html. For instance if jsmith@ucsd.edu and gmac@ucsd.edu worked on the assignment they would call the file behavior_jsmith_gmac.html. Use a txt extension instead if you have written a plain text file.
2. Attach this file to an e-mail message addressed to cg8w1@icogsci1.ucsd.edu, with the subject "Behavior assignment"
3. You must send the message no later than Sunday, February 4 at 11:59pm.