Ultrasonic Range?
 

Hi,
My team is looking into using an ultrasonic sensor and gyro sensor combination for aiming. And we want to be able to shoot from 40 feet away but the current sensor we have (XL‑MaxSonar‑EZ0) only senses up to 25 feet. Are there any other sensors out there that go farther? Or can we use the M1011 axis camera to do that?

Any help would be really appreciated!

Re: Ultrasonic Range?
 

20 feet is really the max range on most sonics.

Re: Ultrasonic Range?
 

You might consider using the camera. Our team has found it to be accurate within thenths of an inch, which is really all you should need. Much of the code can be easily found through the labview examples. Cameras are also beneficial to use during the game in case you need to see more

Re: Ultrasonic Range?
 

If you're 40 feet from the goal at one end of the field, there's a much closer wall at the other end you can sense your distance from.

Re: Ultrasonic Range?
 

THe poster above me makes a good point about measuring your distance for the other side as its closer. We are tackling this problem by using high 1mm resalutin ultrasonic sensors at varios points in the gaem and using a gyro/accelerometer and a Khalman filter + 2 rotary encoders to figure out our translations from there and estimate posistion (acurately) untill conditions are met to reset our absolute posistion with the ultrasonic sensors. Now I dont know if this is pluasable for you to do this late in the season but if interested I am willing to share our algorithms with you.

Re: Ultrasonic Range?
 

Wow! That sounds like you're doing full blown sense-plan-act navigation. I'd love to see how you're doing it, even if you just post your code after the season.

Do you handle situations where the ultrasound or gyro or encoders drop out? One of the strengths of a Kalman filter is making predictions with incomplete and irregular information.

Re: Ultrasonic Range?
 

Is a Khalman filter similar/superior to a moving average?

Re: Ultrasonic Range?
 

Kalman, no "h".

It's more like a hidden Markov model than a moving average. I first heard of them in an AI course--there are several excellent, free, on-line courses that cover them. My only experience with them was that course and they are much more complicated than I'd be comfortable using on an FRC team. I'm really curious to see Jcbconway's implementation.