Why use PING (ultrasonic) sensors instead if IR sensors, or vice versa?

JimG

What advantage is there to using PING sensors, other than distance, rahter than IR sensors? Or why use IR instead of PING.


Thanks,

Jim

Duane Degn

I added an ultrasound sensor to my Scribbler 2 robot to enhance its obstacle detection ability.

The S2 comes with IR sensors, the ultrasound plus IR are better at detecting obstacles than just using one kind of sensor.

I explain a bit about the advantage in the video and thread.

Note: A Ping sensor will not easily work with the S2.

JimG

Duane,

Thanks for the reply. I understand that the idea is to enhance obstacle detection, I am just trying to understand how or why it enhances it.

Jim

Beau Schwabe

Both types of sensors are susceptible to external interference, external light sources, external sounds , etc, At any give time it is less likely that both sensors are compromised by noise at the same time. In a case where only one sensor might be affected, the collaboration of another type of sensor can help 'fill in' any gaps.

erco

Sensor fusion, AKA sensor confusion.

JimG

I was looking at adding IR to a robot I have that alrady has PINGs around it to improve the error rate but, how do you know which one is correct? If you get mixed messages from a PING and an IR how do you determine which value to act on?

Duane Degn

In the video I linked to, IIRC I describe how soft items aren't always seen by ultrasound.

I've found narrow furniture legs aren't seen by the IR sensors but the ultrasound sensor does see them.

If you get conflicting information from the two sensors, it's usually safest to assume the one reading an obstacle is correct.

Edit: I just re-watched the video. It doesn't give a good example of when IR would have been better than ultrasound. If the couch had continued to the floor, I think the IR sensors would have seen it when the ultrasound couldn't.

When I've used my S2 in the kitchen, with lots of hard flat surfaces, the S2 navigates with IR more than ultrasound since the ultrasound doesn't echo back if it hits the hard flat surface at too great an angle.

BTW, sorry for such a short initial answer. I thought more of this was discussed in my S2 thread, but now I realize these discussions had occured in other threads.

JimG

Duane,

Now I understand where you were headed.

I am debating about replacing my present PING mounts with the dual sensor mounts that hold a PING and an IR sensor in one unit. I am trying to determine whether it would be better to have the two different sensors side by side or one above the other as in the dual sensor mount...any thoughts?

Jim

Duane Degn

Jim,

I don't have enough experience with these sensors to give you a good answer.

I'm pretty sure a lot will depend on the objects you wish to detect.

To make the ultrasound demo video, I thought it would be good to have a smaller space so the S2 would run into objects more frequently. I started setting up a bunch of cardboard walls to enclose an area. I found the ultrasound wasn't needed in my test area since the IR sensors worked so well on the light colored cardboard.

As you saw, I ended up just using our livingroom space for the test since it has more obstacles better seen by ultrasound.

So depending on where your robot will be wandering, you might not need one of the sensors at all. It will likely take some experimentation to find the best setup for your environment.