Recent Changes - Search:

TEAMS Academy Wiki

Explore TEAMS!
for visiting sophomores & juniors

Robotics

EnvBioTech

Bat Design

Assistive Tech


Students


Instructors

TEAMS Forum

TEAMS Calendar

TEAMS Web Site

Wiki Info

edit Student.SideBar

GeorgeAndChrisToTheRescue


Plan of Action:
So as the title indicates, Chris and I plan on creating a robot that will navigate the maze to find the stranded person (blinking light). The the idea behind it is focusing on the white walls, and having the robot respond to changes when the sensor identifies the black tape. However, we need the robot to turn properly and identify the blinking light, which comes down to simple guess and test, how powerful the motors should be, how long should they run, etc.

First Plan:
The first plan was simple, just move the robot in a straight line, because the DC motors were acting wierd we needed to find out what levels they could run at to spin the wheels at the same rate. After that adventure we planned to have a single light sensor on the right hand side of the robot that when sensed a change would shut off the corresponding right wheel. Soon we discovered that such a route did not work, because the robot would endlessly bump into things. Thus the idea of two light sensors was proposed, so that when either one deviated from white light, that corresponding sides motor would shut of for a specifies amount of time, allowing the cricket to turn at a broad curvature (some fraction of -x^2). Thus through another phase of guessing and testing we will be able to figure out the times necesarry for these turns( although there is probably some mathematical relationship between the power of the motors, the gear reduction, torque, friction, and the general curvature to determine the time). In this way, the robot always has a bearng of what is immediately to its left and right and can respond to it.

Problems:
Some serious problems we faced were getting both wheels to thouch the ground (we were missing an egg) and to get both wheels to turn at generally the same rate to produce a straight line. Another problem not quite considered yet is the immediate front of the robot is invisible to it. Thus by perhaps adding a touch sensor, so that if it senses hitting a wall it would back up for x amount of time then try to get its bearings by turning either left or right. Now, to sense the blinking light, the robot will have to have a fourth sensor, telling it that light is changing nearby, in which case it should stop, if the program is aligned correctly. For this feature we might include a command instructing the robot to continue towards the light and when it bumps into something, here a second touch sensor could be necesary, it will back up slightly and stop.

Edit - History - Print - Recent Changes - Search
Page last modified on September 13, 2008, at 02:29 PM