Designing Behaviors: Living Robot Simulation

The Living Robot

DESCRIPTION

The Living Robot is a simple simulation of a robot that uses seven people linked together in a group, each mimicking a different component of the robot's behavior by following a precise set of instructions. This simulation aims to bring some of the basic principles of software engineering to life and give non-technical people a feeling for what is involved in designing software to control robots.

When the independent instructions of each component are followed properly and precisely, the robot (a composite of the seven people) performs the desired actions. In this way, the participants can comprehend the resulting behavior of the whole robot as an integrated set of interdependent smaller actions.

Here is a schematic of the robot's configuration:

COMPONENTS

Here are the functions of the seven components:

BRAIN: controls all operations and actions of the other components
EYES: sense the intensity of the ambient light in front
LEFT ARM: senses physical contact with an object on the left side
RIGHT ARM: senses physical contact with an object on the right side
LEFT LEG: propels the left side of the robot forward or backward
RIGHT LEG: propels the right side of the robot forward or backward
TAIL: signals with rear-facing lamp

Using this configuration, we specify a behavior as follows:

The robot goes steadily forward until a light is shined on it, when it stops. While moving, if it is bumped on the right side, it will turn to the left and turn on the tail lamp during the turn. Similarly, if bumped on the left, it will turn to the right with the lamp indicating this.

In order to accomplish this, the brain must synchronize the activities of all of the other components. It continuously monitors the information arriving from the input sensors–the eyes, left arm, and right arm. Processing that information, it then directs the activities of the action components–the left leg, right leg, and tail lamp.

INSTRUCTIONS

Here are the instructions for each component of the Living Robot:

Component	WHEN YOU...	DO THIS...
LEFT ARM	contact anything	say LEFT CONTACT
RIGHT ARM	contact anything	say RIGHT CONTACT
EYES	see a bright light	say BRIGHT LIGHT
LEFT LEG	hear LEFT LEG GO hear LEFT LEG STOP	walk slowly forward stop walking
RIGHT LEG	hear RIGHT LEG GO hear RIGHT LEG STOP	walk slowly forward stop walking
TAIL	hear LAMP ON hear LAMP OFF	turn on light turn off light

The brain is a little more complicated because it has to orchestrate many actions based on input from the other components:

Component	WHEN YOU...	DO THIS...
BRAIN	want to start the robot	say LAMP OFF say LEFT LEG GO say RIGHT LEG GO
	hear BRIGHT LIGHT	say LAMP OFF say LEFT LEG STOP say RIGHT LEG STOP
	hear LEFT CONTACT	say LAMP ON say RIGHT LEG STOP wait one second say RIGHT LEG GO say LAMP OFF
	hear RIGHT CONTACT	say LAMP ON say LEFT LEG STOP wait one second say LEFT LEG GO say LAMP OFF

A printer-friendly version of these instructions can be found here.

IMPLEMENTATION

For the microprocessor version of the simulation, a LEGO Mindstorms robot is constructed and a software program written to implement the behavior. The robot has 2 touch sensors (for the left and right arms) and a light sensor representing the eyes. Two motors are used for the left and right legs, and a simple bulb is used for the tail lamp.

Here are some photos of an implementation on an older-style RCX robot that implements this design:

You can click on any of these to see a large version in separate window.

Here is a picture of a Robolab program implements this design:

You can click on this to see a large version in separate window.

And finally, you can find the actual Robolab program here.

OK, I'm lying, for now. This will be an assignment later in the course, so I'm not putting them online until after you do it.