My First Robot - A Differential Drive Dummy Robot (Version 2)

[reefat]

Hi everyone, this is the improved version of my previous Differential Drive Robot. I didn’t think about any name yet; but will figure out one soon. Anyway, this improved prototype has an IR Rangefinder attached to it which can detect distance of any object in front of it within a range of 10cm to 80cm. And an extra servo is added to this new version to rotate the sensor left or right. Let’s take a closer look at the new prototype.



Whenever the robot faces an obstacle in front of it, it looks left and right, and thinks about which side has enough space to go for. Then it turns to that side and moves forward. Check out the following video to see how the robot moves.

[youtube]s9aAY5zDuLo[/youtube]

[pomprocker]

reefat, your servo and sharp IR mount look just like mine. Note that the $50 sharp IR code comes from Admin's sumo robot. Look at pictures of that robot and youll see you have your sharp IR mounted backward from the way he has his. this may or may not make a difference i'm still not sure.

[reefat]

Thanks for your info. I mounted the IR wrong. I'll fix it. 
And, the code I wrote myself. I have to study and work on the Sharp IR a lot. First I took 100 data (distance vs. voltage) and plot them on graph. Then using MATLAB I had to derive an equation (using Regression) that best-fits with my graph. Although this is a lot of work to invent "THE WHEEL", which is already invented. I like to build everything from scratch. 

[Webbot]

Hay Reefat,

So it looks like your robot goes in a straight line until there is a wall directly in front. You then stop: scan left/right to work out which way to go and then go in a straight line again.

Why not scan your sensor left/right whilst you are going in a straight line? Then you can start to turn away from walls BEFORE you get too close. Make your turn angle inversely proportional to how close the wall is. ie if the wall is far away then turn away slowly. If its very close then turn quickly. Then you will have a much more fluid movement.

The next challenge for you is small objects like chair legs, or low objects like the horizontal struts on a table leg. They may be below the line of site of your Sharp IR. How will you handle those?

Dont mean to sound critical - just giving you more challenges to think about!

[reefat]

@Webbot. I am so glad to have some gr8 !dea and challenges from an expert like you. I appreciate it. I will work on those and let you know.
Right now trying to make a good and stable structure (mechanical body construction). Hope, sooner later I can finish it.

[pomprocker]

Hay Reefat,

So it looks like your robot goes in a straight line until there is a wall directly in front. You then stop: scan left/right to work out which way to go and then go in a straight line again.

Why not scan your sensor left/right whilst you are going in a straight line? Then you can start to turn away from walls BEFORE you get too close. Make your turn angle inversely proportional to how close the wall is. ie if the wall is far away then turn away slowly. If its very close then turn quickly. Then you will have a much more fluid movement.

The next challenge for you is small objects like chair legs, or low objects like the horizontal struts on a table leg. They may be below the line of site of your Sharp IR. How will you handle those?

Dont mean to sound critical - just giving you more challenges to think about!

Thats where I'm at, i've seen my robot hit chair legs, and not see objects below the scanner. Although I have a touchless bumper mounter lower then the scanner. I figure any lower than that and the robot can run it over =P

[eduard2000]

 I think you could thy to put a little more wait to the back because if you would want it to go faster it would tern over and fall 
 But seal nice robot

[GalacticNerd]

I'm making one just like it =)
This is all pretty new to me so.. I hope someone can help me if I'm stuck^^

[Kirk]

I LOVE the cardboard box chassis!  So many builders get all concerned about making the perfect chassis and never get the robot running.  You on the other hand have a cool robot. Good work.  After seeing your robot I think my next robot will use a wood cigar box for a chassis.
Keep posting
Kirk