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 / 25630
25629  |  25631
Subject: 
Re: turtle coding kit (Re: How would we (the rest of us) communicate ideas to the MDP?)
Newsgroups: 
lugnet.robotics
Date: 
Wed, 22 Feb 2006 19:26:11 GMT
Viewed: 
2168 times
  
In lugnet.robotics, Justin Fisher wrote:

<snip>

I'd go further than that - I may be wrong here, but AFAIK professional
roboticists largely consider dead reckoning to be a losing game, an amateur
red-herring that you can waste all the time and money you want on and it will
still never work reliably - traction differs as the robot travels various
surfaces, wheels slip, a pet or collision or other interaction moves the robot
without it being aware, a gear slips, etc etc. Watching a motor counter instead
of the world seems to be the one way to guarentee that a robot never knows where
it is, where to go, or how to navigate the obstacles around it.

I don't know anything about the FLL competitions, but it sounds like they're
designed to offer highly consistant conditions in order to make dead-reckoning a
plausible design approach. I wonder if this might work slightly against learning
about making real-world useful robots? Motor counters are incredibly useful in
all sorts of ways for robots and other Mindstorms things, but if people are
excited over their potential applications for navigation, that seems like
something isn't right.

But like I said, I'm not familiar with how the FLL comps work.

When I was a mentor for a local FLL team, I was so hard pressed to convince the
kids that 'dead reconing' wasn't the way to go.

THey had the entire program coded like 'go straight for 10 seconds, then turn
left for 2', thinking that this would get them trhu the maze.

THe problem was, as I kept trying to show them, is that if you start your 'bot
even at the slightest different position or angle, your 'bot will be so far out
of position by the time it should have reached the finish line in the maze.

But they persisted in the timing solution.  didn't do so well...

Eh, it was a valiant effort.

Dave K



Message has 1 Reply:
  Re: turtle coding kit (Re: How would we (the rest of us) communicate ideas to the MDP?)
 
(...) This sounds like a good demo that someone within the MDP program can create (hint, hint, wink wink, nudge, nudge). Show the difference between navigating through a maze using (1) the 'dead reconing' method, (2) the ultrasound sensor and the (...) (19 years ago, 22-Feb-06, to lugnet.robotics)

Message is in Reply To:
  Re: turtle coding kit (Re: How would we (the rest of us) communicate ideas to the MDP?)
 
(...) I'd go further than that - I may be wrong here, but AFAIK professional roboticists largely consider dead reckoning to be a losing game, an amateur red-herring that you can waste all the time and money you want on and it will still never work (...) (19 years ago, 22-Feb-06, to lugnet.robotics)

18 Messages in This Thread:







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

This Message and its Replies on One Page:
Nested:  All | Brief | Compact | Dots
Linear:  All | Brief | Compact
    

Custom Search

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