CBLL HOME
VLG Group
News/Events
Seminars
People
Research
Publications
Talks
Demos
Datasets
Software
Courses
Links
Group Meetings
Join CBLL
Y. LeCun's website
CS at Courant
Courant Institute
NYU
Lush
Lush

Introduction to Robotics: Material and Pointers


[ Course Homepage | Schedule and Course Material | Videos/Photos | Mailing List ]

Assignment 01: Arduino, Sensors, and Servos

The NYU computer store has received about 100 Arduino boards on 01/23. Get yours before Tuesday!

In addition to the Arduino, you will need the following parts (one set of part for two students is OK).

You can either get the Arduino kit box (but the store doesn't have many of them left), or you can buy the necessary items spearately from the computer store:

You also need resistors, wires, and a small DC motor. You can get assorted resistors from the computer store and wires from Radio Shack (two block north of the classroom on Broadway), but you can also use one from the lab. The equipment is set up on a set of tables in the south-west corner of the 12th floor.

The TA is Pierre Sermanet [pierre.sermanet [at] gmail [dot] com]. Koray Kavukcuoglu [koray [at] cs [dot] nyu [dot] edu] will also be present in the class on 01/27 and 01/29.

The projects for 01/27 and 01/29 are:

  • reading analog values from the CDS cells (connecting a resistor as a voltage divider)
  • reading analog values from the flex sensors (connecting a resistor as a voltage divider)
  • controlling a servo: you must generate a pulse between 1.0 and 2.0 millisecond every 50 millisecond to control the servo. The width of the pulse determines the positon of the servo arm.
  • control the servo position using the flex sensor.
  • connect 2 or 3 CDS cells to the arduino, stick the prototyping breadboard on the servo arm (using double-sided tape), and control the servo to orient the board in the direction of highest light level.
  • Extra credit: Implement a proportional controller with dead band, i.e. the speed with which the servo is rotated should be proportional to the "error" signal. The error is the difference between the ouputs of the left and right light sensors (sensors should probably be calibrated during the startup phase).

Assignment 02: PWM speed control

Control the speed of a DC motor using Pulse Width Modulation.

controlling an electric DC motor: connect the H-bridge chip and control the speed of the motor using Pulse-Width Modulation using the AnalogWrite() function of the Arduino.

Assignment 03: following a line with the 3pi robot

3PI robot: calibrating the reflective light sensors.

Get this sample code that reads the line sensors

3PI robot: line following with proportional controller. Extra credit for implementing a PID controler. Super extra credit for winning the race.

Assignment 04: 3pi dead reckoning

3PI robot: dead reckoning and driving home. The robot must follow a line, detect the end of the line, and return to the starting point (or as close to it as possible).

You must write a function with the following prototype:
"void update_position(int ml, int mr, long dt, long *x, long *y, long *theta)"
where ml and mr are the motor speeds, dt is the time interval during which these speeds are valid (in milliseconds), and x, y, and theta are pointers to the coordinates and heading variables of the robot (to be updated by the function.

you will need trig functions as well as functions to convert the motor speeds to robot speed and rotation.

To help you with this, we are providing the following .h file, which contains trig functions, as well as functions to convert 3pi motor speeds to linear and rotational speeds: 3pi_kinematics.h.

Add the file to your directory, and add '#include "3pi_kinematics.h"' to your program.

The file provides four functions:

  • long Sin(long a): return 1000 times the sin of angle a (in degrees).
  • long Cos(ling a): return 1000 times the cos of angle a (in degrees).
  • long motor2speed(int v): return the speed of the robot in 1/10 mm/s given the average motor speed of the two wheels.
  • long motor2angle(int mleft, int mright): return the rotational speed of the robot in degrees per second, given the motor speeds of the left and right wheels.

Note that the speed and rotation functions have been calibrated on a particular robot. Your own robot may have slightly different characteristics. Also note that the estimates are not likely to be accurate if the speed changes abruptly.

Assignment 05

Detect a tennis ball and drive the rovio to it, stopping a few cm before it.

You need Lush, which runs on Linux and Mac.

You also need the Lush Rovio library.

First, you need to estimate the distance from the rovio to the ball.

Using the diagram below, find the formula that relates gives x from L, where x is the distance to the ball, and L is the difference between row number of the bottom of the ball in the image, and the middle row of the image.

Next, you need to calibrate the distance driven by the rovio when it receives driving commands.

The rovio drives for a short time for each command it receives. If you don't keep sending driving commands at a fast rate, it will stop. If you want to stop it right away, send it a stop command. See the rovio-lib.lsh file for examples.

Assignment 06

Get your rovio to play soccer!

The goals are bright red. You must be able to detect the goal so you can drive between the goal posts.

The strategy is to turn around to find the ball, orbit around the ball until the goal is visible, and drive towards the goal.

.