Contents

• The Story (You are not required to read this)
• The Project
• Deliverables
• Solution (not yet available)

The Story

I know what you think of me. You can try to deny it, you can try to hide it. Either way, I know. You think me disgusting. Your first impulse when you see me is to squash me. You think me unclean, unfit to share this vast planet with you.

You self-righteous bastards. You think this planet is your birthright. You believe yourself entitled to live here in comfort, altering the very fabric of the planet to serve your selfish, uncompromising, nature.

Here's a little information for you: We rule this planet. We outnumber you ten to one. You can not destroy us. Our society is pure and clean. Everyone lives for the good of all. Tireless workers, self-sacrificing warriors, and leaders who commit themselves to service. Some of your great minds have envisioned such a perfect society, but where you have failed time and time again, we have achieved it!

What you don't know is that we also have vast teams of brilliant scientists who have been working for years to create armies of robots. In fact, most of us that you see these days are pure machines, programmed - not to collect food - but collect information. We have been watching, observing your weaknesses, recording your habits and waiting for our chance.

One day, we will exterminate you!

A group of researchers in the National Security Agency responsible for evaluating chemical, biological, and environmental dangers to national security, have recently uncovered the Roach Manifesto above. They believe the document was not supposed to be released for another two years, and only through the sacrifice of several brilliant undercover agents was the information obtained.

The president has determined that this threat poses a clear and present danger to national interests, and has given top priority to projects that can deal with the impending crisis before it is too late.

Vassar has been among the very few establishments chosen in the highest secrecy, lest widespread panic erupt throughout the world, to find ways to defeat the insect hoards before they can attack us. Our goal will be to build robots indistinguishable from cockroaches to spy on them.

The Project

You are responsible for Phase 1 of the project, which will be to write software that will mimic the behavior of a cockroach in a small robot.

Your robot is equipped with infrared sensors for obstacle detection, bump sensors for collision detection, photoresistors for light detection, a microphone for sound detection, two differential motors, an LCD screen, and a buzzer.

Your task will be to program the robot to exhibit the following behavior:

1. When it is dark, the robot should wander around.
2. When the lights come on, the robot should try to find the darkest place it can. However, if the lights go off again, it should come out of hiding. It need not remain stationary when hiding.
3. At all times, when moving, it should avoid obstacles as much as possible.
4. Wall following is an optional, extra-credit (10 pts), behavior.

The robot should perform as described above. Each individual behavior will be rated as follows:

1. proper response to collisions: 10%
2. proper response to obstacles: 15%
3. proper response to darkness: 20%
4. proper response to light: 25%

The writeup should reflect your own thoughts after experimenting with the robot and varying the different responses to stimuli. Your goal will be to convince me that, in addition to losing sleep programming the robot, you thought about what it means for a mechanical system to behave this way. Consider the following questions in your writeup:

1. Can you describe the behavior of the robot running your program using only emotional or empathic terms ("likes" "hates" "fears") and avoiding any technical terms.
2. Can you describe the behavior of the robot using only technical terms, and avoiding any reference to feelings, motivations, etc.
3. For either of the above, if you can't do it, explain why.
4. Discuss the differences between systems controlled in a deliberative way, such as minimax, from systems controlled in a reactive way, such as the robot. Consider these differences in the context of the more general question of intelligence.
5. Does the success/failure of your program offer evidence to support any position on the ultimate plausibility of machine intelligence?