Outdoor step stone pressure/presence sensor

I'm creating a project I call "Talking Stone" and need to figure out a good rugged pressure or presence sensor for them. Basically, I'm going to cast cement stones for my backyard and I want to have some sort of a sensor to detect when a stone has been stepped on. Any ideas on what type sensor I should use? It needs to be something that will be rugged, last a long time and not get stuck "on" by the weight of the stone.

More info on the project:
For the casual viewer the stones will look like standard walking stones with a few clear marbles on the surface. When someone steps on the stone, LEDs in the "marbles" (cast out of a polyester resin) will light up and fade out. Inside the stone will be a radio transceiver that communicates to a base station. The base station will control how the stones react to being stepped on. It will have several programs that might allow some stones to react to others or have them all light up in sync to some music.

Thanks!

I'm wondering if a Piezo sensor would work, like the Guitar pickup seen here: Piezoelectric sensor - Wikipedia

A piezo would probably work. Then the question is how to mount it. The stone will have some weight, so it will already be pressing down on the element. Like a switch the stone will have to reliably turn "on" when stepped on and "off" when stepped off. What would be a good way to set that up? A special platform? Another thing to consider is that the stone will probably be surrounded by dirt or sand.

How about instead of placing the sensor under the stone you place it inside the stone? If your casting the stone you could use pvc and place it in the center of the stone which would house your sensor. This would allow it to not be crushed by the weight of the stone and the weight of the person. Then to get creative you could even put plexiglass over the sensor and add some leds for fun. Run the wires out the bottom and presto. Just a thought..

I would try a Force Sensitive Resistor.

microdevil45:
How about instead of placing the sensor under the stone you place it inside the stone? If your casting the stone you could use pvc and place it in the center of the stone which would house your sensor. This would allow it to not be crushed by the weight of the stone and the weight of the person. Then to get creative you could even put plexiglass over the sensor and add some leds for fun. Run the wires out the bottom and presto. Just a thought..

Good idea. How would the sensor in the PVC be activated? Would it be sticking out the top or bottom?

Pauly:
I would try a Force Sensitive Resistor.
Force Sensitive Resistor - Square - SEN-09376 - SparkFun Electronics

That's cool. How/where would you mount it to make sure the weight of the stone wouldn't keep it turned "on" and protect it from being destroyed by frequent use? Thanks.

A other option is:
Air pressure. Under the stone you layout a air-hose. You close one end and mount a pressure sensor on the other end. This is almost unbreakable and you can ruffly measure the weight of the person that is stepping the stone.

Depending on the mass of the stone and placement of the sensor you might be able to use an accelerometer to pick up vibration from stepping on the stone. If the accelerometer can successfully pick up the vibration then you could use a digital accelerometer that has an embedded "single tap" algorithm (like the ADXL345). It would cut down on your code because it would be interrupt driven.

Pressure sensitive resistor holds some promise. Expensive if you have qty needed. It would also require a firm foundation to press against. The two surfaces would have to be slightly raised and smooth for it to work well.

Pressure sensitive resistor sounds workable,

BUT whatever you do there will be drift and change, so a fixed calibration may be a problem.

I think you want your code to keep calculating a running average reading and then look for a sudden change > some threshold. That
would be much more reliable at sensing the step-on.

I'd prototype this and save and graph a bunch of data to characterize the response. Then decide your final algorithm.

Let us know what you end up with!

Personally, I would cast some wire into the top bit of the stepping stone and use a capacitance sensor with some auto-sensitivity code.
Someone's foot would change the capacitance quickly, while weather conditions and such would change it much more slowly.
When the arduino detects a sudden change, it fires the lights (a note: The lights will almost certainly pin the cap. sensor at full, turn off the auto-sensitivity code while the lights are on!).
In the following code the cap sense bits have been hacked out to just leave the (very simple) auto sensitivity code, the variable "fout" is the output from the capsense code.
Change the interval variable to change how fast the capacitance change has to be to set the lights off.

if (millis() - previousMillis > interval){
   old4 = old3;
   old3 = old2;
   old2 = old1;
   old1 = fout;
   recent = (old1 + old2 + old3 + old4) / 4;
   previousMillis = millis();
 }
 
 if (fout - recent > 10){
 digitalWrite(3, HIGH);
delay(1000);
digitalWrite(3, LOW);
 }

A lot of great ideas so far! Keep them coming.

The air pressure idea seems to have quite a bit of promise; due to how rugged it would be. What type of a pressure sensor would you use? Looks like the barometric sensor from Sparkfun (Barometric Pressure Sensor - BMP085 Breakout - SEN-09694 - SparkFun Electronics) could work, but I would prefer a cheaper option if possible. Are there good, cheap, air pressure sensors somewhere? I'll probably be making between 5 to 10 (or more) stones, so I'm trying to keep the cost of parts down. Then again, there is value in building something that wont break after a couple months.

Cinezaster:
A other option is:
Air pressure. Under the stone you layout a air-hose. You close one end and mount a pressure sensor on the other end. This is almost unbreakable and you can ruffly measure the weight of the person that is stepping the stone.

WOW! it is because of threads like this i love this forum, taking part of different peoples often brilliant ideas, it feels great.
However i'll put my two cents in here to.

Force or pressure? i think both will need a solid foundation. Force sensors, i think you will need to have one in each corner of every tile, expensive and i actually dont think they will stand up against the elements of nature, rain, snow etc. Not the ones linked to anyway.

I made a 3D-"sketch" of the pressure idea wich was a very brilliant idea.

tile3 by swetrack, on Flickr
As you see there r some things you have to consider, fixate the tubing, dirt may fall down between the tiles, the sensor and hose clamp needs to fit somewere, the tiles might be swampy to walk on and so on but i am certain that it is possible.

Another idea was a capacitive sensor: well... i dont think so, dirt rain and wear will make it difficult.
And there was the accelerometer idea, wich is very very good, it might need some analyzis of materials, mountings, soil-damping and so on it was so interesting it made me think of a similar idea but with a michrophone instead.

tilemic by swetrack, on Flickr
Think of it, when you put a glass to a wall and your ear to the bottom of it, it amplifies the smallest taps to almost painfully loud noise. I thought you could make it similar with the tiles but maybe with a can or so. In the bottom of the can you put a small michrophone. It will be easy to make a hole in the soil so the can and mic will fit. It will take some testing first and probarly lots of software debugging, but hey, thats half the fun.
It will be like AC-coupled though, people needs to move to make it light and rain will probarly trigger it to. But maybe that would be cool?

I did not mean to step on anyones toes here, this is just my thoughts about it.
And sorry about flickr spam but dont know any other way to put in pictures here.

thank you

I wonder whether this conductive foam pressure sensor hack would work for your application:

-br

http://entropymouse.com

The capacitance sensor is a really cool idea and best of all, no moving parts. I haven't worked with capacitance sensors before, have done a bit of googling, but still have a couple questions:

  • How much wire would have to be at the top of the stone? (i.e. would the shoe have to come in contact to trigger or just be close enough to it)
  • Would the sensor pickup capacitance of a rubber shoe?

It would also be cool to have part of the program setup to react to slow changes so we could get a light show in the middle of a rain shower!

Bobnova:
Personally, I would cast some wire into the top bit of the stepping stone and use a capacitance sensor with some auto-sensitivity code.
Someone's foot would change the capacitance quickly, while weather conditions and such would change it much more slowly.
When the arduino detects a sudden change, it fires the lights (a note: The lights will almost certainly pin the cap. sensor at full, turn off the auto-sensitivity code while the lights are on!).
In the following code the cap sense bits have been hacked out to just leave the (very simple) auto sensitivity code, the variable "fout" is the output from the capsense code.
Change the interval variable to change how fast the capacitance change has to be to set the lights off.

if (millis() - previousMillis > interval){

old4 = old3;
  old3 = old2;
  old2 = old1;
  old1 = fout;
  recent = (old1 + old2 + old3 + old4) / 4;
  previousMillis = millis();
}

if (fout - recent > 10){
digitalWrite(3, HIGH);
delay(1000);
digitalWrite(3, LOW);
}

Also, is there a preferred sensor to use for this? Would this one from SparkFun work? http://www.sparkfun.com/products/9695

swetrack:
And there was the accelerometer idea, wich is very very good, it might need some analyzis of materials, mountings, soil-damping and so on it was so interesting it made me think of a similar idea but with a michrophone instead.

tilemic by swetrack, on Flickr
Think of it, when you put a glass to a wall and your ear to the bottom of it, it amplifies the smallest taps to almost painfully loud noise. I thought you could make it similar with the tiles but maybe with a can or so. In the bottom of the can you put a small michrophone. It will be easy to make a hole in the soil so the can and mic will fit. It will take some testing first and probarly lots of software debugging, but hey, thats half the fun.
It will be like AC-coupled though, people needs to move to make it light and rain will probarly trigger it to. But maybe that would be cool?

I did not mean to step on anyones toes here, this is just my thoughts about it.
And sorry about flickr spam but dont know any other way to put in pictures here.

Wow, thanks for creating a couple 3D mocks! The microphone is an interesting idea. I like how it could react to the sound of the rain falling. It would have to be calibrated to not fire at any loud sound.

That has quite a lot of promise! I wonder how long those foam things will last before they loose their "bounce".

billroy:
I wonder whether this conductive foam pressure sensor hack would work for your application:

http://www.instructables.com/id/How-to-Make-a-Ridiculously-Cheap-Analog-Pressure-S/

-br
http://bitlash.net
http://entropymouse.com

Would fluid work?