To LUGNET HomepageTo LUGNET News HomepageTo LUGNET Guide Homepage
 Help on Searching
 
Post new message to lugnet.roboticsOpen lugnet.robotics in your NNTP NewsreaderTo LUGNET News Traffic PageSign In (Members)
 Robotics / 3450
3449  |  3451
Subject: 
Re: What is your current Mindstorm project?
Newsgroups: 
lugnet.robotics
Date: 
Thu, 4 Feb 1999 16:25:51 GMT
Viewed: 
1428 times
  
Matthew Miller writes:
P.S.  So what projects are you working on with your Mindstorm?

Two ideas.

[ ... ]
2) A room-mapping robot that communicates with a Linux computer which does
   the actual storing of the map and computing where to move next.

This is exactly what I'm working on (as a long term goal). I only got
my mindstorm a few days ago so I haven't gotten far. Currently I'm
just trying to get my robot to survive in my flat for more than a
minute! ;-)

The main problems I see, apart from a basic dont-get-stuck
competence, are:

- Position sensing. How do you know where you are well enough
  to construct an accurate map? In particular, how do you cope with,
  for instance, scraping along a non-sensed wall which slightly alters
  your trajectory?

- How do you transmit the information back? That is, do you log it,
  and then get picked up by your owner who reads your log, or do you
  periodically find home again, such that you have a line of sight,
  and then upload your map to the computer?

Pragmatically speaking, I only have a limited amount of bits and a
limited budget, so making a robot which can survive my flat, map
a room *and* transmit its map back to base and then continue, may be a
long shot.

Other issues which I'm thinking about are map representation. My
current idea is to designate gridded areas relative to the initial start
position of the robot. Assuming it knows when it is in one segment of this
grid, it should keep a count of how many times it has hit anything in
that area. Once the count reaches a certain threshold, that grid is
designated as "solid" and will labelled as such. The ids of grid areas
which contain solids would then be transmitted back as the map. Since
my current design of robot has a turn-on-spot/go forward/go backward
ability, it may be better to have a more polar-coordinate path based
system instead.

Any thoughts?

--

Mike Moran, Dept. of Electrical Engineering, The University of Edinburgh
King's Buildings, Mayfield Rd, Edinburgh EH9 3JL



Message is in Reply To:
  Re: What is your current Mindstorm project?
 
(...) Two ideas. 1) My friend Paul and I are trying to make some sort of 'tag' game -- or perhaps hide & seek -- that two RCXbots can play with each other. 2) A room-mapping robot that communicates with a Linux computer which does the actual storing (...) (26 years ago, 30-Sep-98, to lugnet.robotics)

13 Messages in This Thread:





Entire Thread on One Page:
Nested:  All | Brief | Compact | Dots
Linear:  All | Brief | Compact
    

Custom Search

©2005 LUGNET. All rights reserved. - hosted by steinbruch.info GbR