Society of Robots - Robot Forum

Software => Software => Topic started by: anu_2k6 on May 14, 2013, 02:56:09 PM

Title: Maze Hidden Object Search Robot. HELP!
Post by: anu_2k6 on May 14, 2013, 02:56:09 PM
Okay, so basically I am new to this and only just slowly getting my head around programming. I've been looking into some projects and this one interests me. The aim is to ...

Design and construct a Robotic Vehicle that navigates through a complex floor plan of a maze. The path is constructed by using appropriately marked path corridors. A hidden object is placed in of the areas of the maze. The vehicle should navigate through the maze by following the defined path and search for the hidden object and should able to memorize the correct path to the hidden obstacle so that in a second attempt in finding the obstacle should be able to find a direct path to its location.

So .. I kind of have more of an idea about the path memorizing part of the robot, but not so sure on the object searching part. I am obviously not expecting a full coding and answers as I would like to learn something myself. I'm looking to see if someone can lead me in the right direction and help me a little as to where I should start for the object searching part? I dont know how to go about it, where to start, what to do etc?  :-[

I did do some research over it but found nothing related to object searching.
Please .. ! Any help will be more than appreciated.
Sorry about the long message!
Title: Re: Maze Hidden Object Search Robot. HELP!
Post by: jwatte on May 14, 2013, 05:01:16 PM
In pass 1, just explore all of the maze, and if you find the object in a particular location, remember it. You may need the entire maze map to be able to calculate the optimal path for pass 2.

What is the problem with the object? Do you not know how to detect the presence of the object? What are the specs for the object -- size, color, shape, IR/sound beacons, etc? Based on those specs, different sensors would be more or less appropriate. Bascially, your sensors need to do two things:
1) Detect blocked areas (walls)
2) Detect the object.

Doing 1) is simple with IR or ultrasound distance sensors, or if "walls" are really tape on the ground, IR line sensors.
Doing 2) may be simplest with a webcam and some image recognition software.