Let's Make Robots!

Asymmetry on Sharp IR Distance Sensors

Some graphs of the Sharp Sensors that could help you when designing your new robots.

Hi guys, If you have worked with Sharp IR Distance Sensors you may have noticed that they are not symmetric at all. They give high values when an object is placed in the side of the emissor.

Take a closer look at this picture to see the asymmetry. The values were gotten by the ADC of an Arduino Board, so 0 means 0V and 1023 means 5V.

 

 

This asymmetry can be a problem for example if you have designed a sumobot with two Sharp IR Sensors in the front and you want your robot to be able to track your opponent. 

If you have placed the sensors normally this is what your robot will see.

As you can see the robot will see much better something in the left than in the right. 

 

But the solution is very easy, you just have to turn one of the sensors 180º vertically and you will get something like these.

Placing the sensors in this way gives to the robot a much more symmetric readings. Thus it will be very easy to track your opponent.

 

Summary:

 

 

May the force go with you!

 

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

You can still maintain the "eyes" look and do the rotate 90degrees that Maxhirez has suggested if you have 2 of the sensors

 = )

 

The other trouble with your "good" picture is that there is a spot right in the middle where it would appear to be less sensative.

Thanks. Its always good to be aware of these things.

If a robot needs to scan for an object (i.e. have a Sharp IR sensor on a servo that rotates and takes a measurement at every degree) then a vertical mounting of the sensor is a must, or the measurements will be affected exactly by that asymmetry. We tend to use the sensor to give the impression of the robot having eyes, mostly because of the effect it has on women and children, but in that case we just use the sensor to determine if there is an object or an open path in the direction the sensor is "looking". In a Sumo robot, just like you are showing, the interest is to detect the oponent in a large field of view, so placing the sensor horisontally makes more sense. Having the left and right sensors placed like in your last drawing makes sense, although not many people realize that, so thank you for pointing it out and explaining why it works that way!

I love the way you approached the problem.

As Maxhirez pointed out, there's also the option of moving the asymetry in the vertical plane.

Those graphs are great!

I think a lot of the problem you're discovering here is due to the tendency of robot makers to want to make the Sharp sensor look more like an animal's eyes.  Since this is an emitter/detector pair, when mounting the sensor so that it looks "right" the emitter will necessarily be opposite the detector, so the signal read will be off by a little due to the fact that the IR signal travels farther when it's looking one way rather than the other.  A less expensive solution might be to rotate a single sensor 90 degrees.  It doesn't look as cool but it functions about the same except with regard to the suggestion of tracking.

Yes, rotating a sensor 90 degrees would be a very good option!