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 / 6966
6965  |  6967
Subject: 
RE: IRPD revisited
Newsgroups: 
lugnet.robotics
Date: 
Mon, 20 Sep 1999 21:01:09 GMT
Original-From: 
Tilman Sporkert <tilman@activesw.*NoSpam*com>
Viewed: 
732 times
  
Although I think your statement about robots breaking apart when
they hit a
wall is a bit over-dramatic. Lego creations stick together pretty well --
I've built a large number of robots with mechanical bump sensors.

Correct. You need to learn how to build your touch sensors in a smarter way.
The most intuitive way is to have a feeler arm that presses a button when it
touches the wall. The problem with this design is that your feeler has a
very short range of travel, and once pressed, it presses onto the sensor,
and can't move any further. Before you have a chance to put your motor in
reverse, the kinetic energy of the moving robot will get transmitted to the
wall by the feeler and sensor. That puts stress on your robot right there,
and can cause it to disintegrate.

Instead, build your feelers so that they press the button in when not
touching anything. When it touches something, it releases the button. You
can build such a feeler so that it can move an inch or more. That's enough
for your robot to stop before it hits the object itself.

Here's a very simple example:

   XXXXXXX
  X
o
XTTTT

T is the touch sensor, o and X are the feeler beam, o is the pivot point.
The robot moves towards the top. Pull the short end of the feeler arm onto
the touch sensor with a rubber band so that it closes. Now if the long end
of the feeler beam touches an object, it opens the touch sensor
immediately - without transmitting any force beyond what is needed to
overcome the strength of the rubber band. And it can rotate a little
further, giving the robot a chance to slow down before it actually hits the
wall hard.

There are many other designs possible. Using an IRPD is of course a lot
simpler than tinkering with arms, pivot points, and rubber bands. And blind
spots are easier to manage, too.


Tilman



Message is in Reply To:
  Re: IRPD revisited
 
(...) Very nice! Although I think your statement about robots breaking apart when they hit a wall is a bit over-dramatic. Lego creations stick together pretty well -- I've built a large number of robots with mechanical bump sensors. That's not to (...) (25 years ago, 18-Sep-99, to lugnet.robotics)

5 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