Author Options:

Motor Control System for large Robot? Answered

The motor System

So I am working on  building a robot using wheelchair motors. I'm trying to control it via an arduino potentially but I'm not sure how to do this. My plan is to potentially use an 8channel 10 amp relay board connected to the arduino, but I'm not sure if 10 amps would be enough for the wheelchair motors? Also how do I go about actually powering the motors (12v Lead Acid Battery?) and interfacing that with the arduino (can I?). My other problem is that I would like to have at least a couple servos on each of two arms of the bot....so could I potentially control all that with the same arduino or would I need either a second arduino or an arduino mega with motor shield to do that? 

The navigation

My other problem is navigating/control.....
For that I'm thinking of using an ultrasonic ping sensor or two if that's possible, again I don't know if I can fit all these things on an arduino or not. and Lastly to actually control it, I know I should be able to program it autonomously with the arduino but what do you think is the best way if I wanted to be able to manually control the bot as well? 

Thank you so much for any guidance you guys can give on this large project!

11 Replies

rickharrisBest Answer (author)2014-08-26

Assessing your location to a degree that allows for a robot butler that won't crash into people etc is going to be seriously difficult.

Your going to need to develop a system that can navigate round the house to an accuracy better than half the width of the device. Your going to need to learn the plan of the building so you know where you are all the time.

Your going to need some sophisticated collision detection system that can detect objects in the stopping/avoiding distance from slightly above the top of the unit all the way to floor level (not everything is at the same height).

Commercial systems use Video cameras and vision recognition software which is beyond the average programmer and the use of arduino etc.

In a factory you would use a wire guide under the floor - Mark the pathway and make sure people kept out of the way to minimise these issues.

Your not going to be able to get the accuracy by blind measuring e.g. GPS/Measuring the rotation of each wheel, or Compass and accelerometer.

It is a hugely difficult task to build an operating, safe and useful robot to operate round the house at a reasonable cost.

As I originally said you need to start small, Look at


This is a robot that is aware of it's surroundings but lives in an enclosed area. If you built one of these as a starting point you would learn so much to get yu ready to develop smarter and bigger machines.

Select as Best AnswerUndo Best Answer

rickharris (author)rickharris2014-08-26

PS this robot (Dizzy) is small and really hard to get to work. I know, I built one as a school project, with a group of students. It isn't easy to get it to work properly, although the hardware is very easy to make.

Select as Best AnswerUndo Best Answer

dstech (author)rickharris2014-08-26

First, Thank you so much for all your suggestions! I understand I need to start small. I will take a look at Dizzy. Also, While I hope to eventually have an autonomous obstacle avoiding bot, a teleoperated one is fine (in fact I would rather have it teleoperated than autonomous). As for starting small, I guess it's just that I feel that I can do it, with the right combination of parts and knowledge. I mean, like if my only goal is just a base that I can drive around that's fine. And to me it seems (if I understand correctly) that this can be done with an arduino, and either a relay board or motor controller board right? and then I interface that with the main battery needed to actually power the motors? (again any specifics here would be helpful) but ultimately that doesn't seem that bad....and I can get those things for 150 bucks or so (again not bad at all really)....I just need a few instructions on how exactly to do this cause I'm not completely sure...Thanks again though for the suggestions you have already given. and any more help which you do give. I honestly appreciate it.

Select as Best AnswerUndo Best Answer

rickharris (author)dstech2014-08-28

feeling and doing are 2 different things. By no means am I trying to discourage you but look at it this way:

Your asking very basic but complicated questions, this is a clear indication of your lack of experience and skill and knowledge. Fine but someone else offering piece meal input isn't the way to go or to gain that knowledge and skill.

You need hands on experience - To get that you may need to go down several different routes - This will cost and it is FAR better to do it on a small scale to gain that experience then to pay out 100's $ on thins you may or may not need, may or may not be able to use.

Once you understand how to make it work at a small level - It becomes a much easier problem to scale up AND you have skills and knowledge that allow you to do that.

Start with Dizzy and follow their build because it works - They already did it. - You may use a different micro controller but the basics just follow. They did it that way for a reason.

Then try to improve it. If your now thinking radio control (teli-operation) then you need to collect information on radio control system and FVP and again build small.


Has a load of information basic and complicated on RC control and FVP - YES they build aircraft but the principles are the same for a ground vehicle.

In a way your really asking others to do your design work for you , perhaps not deliberately but that is what your lack of knowledge amounts to - this isn't going to work even if one of us is willing to do that - First time you get a fault your not going to be able to find it because you didn't design it.

START Small. Your progress may be very fast and in 6 months you may have your robo butler up and running OR you may have decided it isn't going to work. OR you may be on a different track.

Select as Best AnswerUndo Best Answer

dstech (author)rickharris2014-08-28

Thank you very much for your help and your suggestions. I agree with you completely that it's better to mess up and learn on a smaller and cheaper scale for sure. I apologize also if I came across as arrogant or rude or anything. I'm just getting into it. My plan for now is to build sort of a classic arduino bot with tamiya gearbox/motors, a castor wheel, a motor shield and maybe an ultrasonic sensor. We'll see how it goes and move on from there....

Select as Best AnswerUndo Best Answer

iceng (author)2014-08-27

For a low cost intro to robotics you should explore a local Hacker-Space in your area see pointer http://www.wirelessdesignmag.com/news/2014/08/hack...

These organizations will let you do hands-on working experience that you need in lieu of an technical degree as demonstrated by your primitive electrical interfacing requests.

BTW since you don't reveal your city, I could not locate a HACKER-SPACE near you ???

Select as Best AnswerUndo Best Answer

iceng (author)2014-08-25

I would use several Stamps


on segregated sub systems and a GPS module for additional guidance.

Yes its true that there is a randomizing input for civilian GPS use that limits accuracy, but you can increase position resolution by ten if you average 100 readings.

Select as Best AnswerUndo Best Answer

steveastrouk (author)iceng2014-08-26

Selective availability was turned off in 1992, so the precision of GPS is what it is. You can improve the accuracy by averaging, but not the precision.

Survey grade GPS is another thing entirely.....

Select as Best AnswerUndo Best Answer

iceng (author)steveastrouk2014-08-27

While I never said "precision" it is still my understanding that military GPS resolves less then 6 inches.....

Select as Best AnswerUndo Best Answer

The MadScientist (author)2014-08-26

For all the drive systems, you have everything you need from the wheelchairs you get the motors from, they have batteries and speed controllers. So what you might want to be looking at is how to use the audrino to manipulate the wheelchair speed controllers. However it depends on what kind of job you want to do on this robot on whether you want to use that roughish method.

Select as Best AnswerUndo Best Answer

rickharris (author)2014-08-26

Assuming you in the USA The Picaxe system is MUCH cheaper then the Stamp. The Picaxe PIC controllers will communicate with each other so you could make a distributed system with several units talking to a master "brain"

I have posted the like to these manuals many times.




Although for the Picaxe system the principles apply to ALL pic type micro controller systems.

Select as Best AnswerUndo Best Answer