Re: Programming a Local GPS Navigation System into a Mobile Robot

circlechess wrote:

Hello World,

I am in a college robotics club and I would like to make a robot that
can navigate locally (like in one corner of a room to another or from
one room to another, for starters). It would be awesome if it could
respond to voice commands or simple inputs like "office" or "commons"
and be able to know where to go. I would like to be able to program a
"mini map" of the environment or create some way of the robot moving
around it's environment and "learning" where things are located
(perhaps via sensors and memory of location points?).

I am not exactly sure where to start or how realistic this is. I
think Rodney Brooks at the MIT AI Lab has worked on similar things
using mobile and walking robots. I suppose this might be similar, but
I am just not sure where to start in getting this going and what the
prerequisite knowledge to undertaking this would be.

I have some programming experience in C++ and will be exploring Scheme/
Lisp in the coming weeks so I am flexible to whatever language is
needed to most efficiently get this going. I would like the language
not to be too limited, however. I currently have a Laptop with Linux
installed onto it that I was planning to use as the robot "brain" and
hook this up to little motors and wheels, etc. on some sort of base/
platform that I can use to program it to move. I would need to get
sensors so that it doesn't bump into things as it is moving about.

I would ideally like it to be able to have a decent map of how to get
here and there while also using the sensors to do obstacle avoidance.

Also-- are there any competitions, national or international, that
this type of robot could be enrolled? It would be nice to have some
type of standard/structure/format to guide me/my club as we try to
build this robot.

Cheers!

-Yama

Look into the FIRST annual competitions of robotics.

Sonar is probably the cheapest way to map a room. GPS won't work inside,
and has too coarse a resolution anyway.

If you mostly want to explore high-level programming, you might consider
just buying a mobile base, otherwise you'll have to work through the
drives, electronics, and controls (which is also fun).

You'll need to decide whether the brains travel with the bot, or control
remotely. Many really good processor boards exist that can run pretty
complex programs, check the AVR AT series.

Mike



--
newsgroups.comcast.net

--
newsgroups.comcast.net
.

Relevant Pages

  • Programming a Local GPS Navigation System into a Mobile Robot
    ... I am in a college robotics club and I would like to make a robot that ... "mini map" of the environment or create some way of the robot moving ... Lisp in the coming weeks so I am flexible to whatever language is ... sensors so that it doesn't bump into things as it is moving about. ...
    (comp.robotics.misc)
  • Re: room recognition
    ... Take a 360 sonar scan and store it ... a "local" map). ... You know how far the robot could have ... to localize, you consider all cells ...
    (comp.robotics.misc)
  • Re: Associative memory for navigation
    ... and in doing so build an associative memory map of the ... Humans can do it, so it's possible without a doubt. ... robot, for example, to "go to room x" (assuming somehow the robot had ... the robot first has to decode that raw 2D data into a description of a 3D ...
    (comp.robotics.misc)
  • Re: Temporal Learning
    ... >> many degress of freedom and many types of sensors. ... >papers, and they didn't even having any learning mechanisms whatsoever, ... IMO, nothing can emerge without some form of learning. ... IIRC, the robot ...
    (comp.ai.philosophy)
  • Re: What is localization ?
    ... 1)the robot explores your house makes a map ... 2)you turn the robot off, pick him up, and move him to ... 3)the robot gets turned on again, and told to localize. ... and compares the local map ...
    (comp.robotics.misc)