ideas for pressure sensing device?

Hi there,
strictly speaking this isn't directly about robotics but I know that robotics people are the ones most likely to have ideas about this, and
you might find it interesting too. I'm looking for a way to measure how much pressure an infant is putting on a squeezy object that is put into its hand (I'm a researcher in child development), and have the measurement input into a computer so I can record the pressure time series.
I'm not that bothered about what form the object takes so long as it is easily squeezed (although it doesn't have to deform a lot it should deform a bit) and it can fit in an infant hand. It's seems to me that the best way to do this is to use an air bladder with a pneumatic pressure sensor.
Now what I am hoping is that such a device might already exist, available off the shelf as it were. Anyone know? If not, I suppose it shouldn't be too hard to make one up from an air bladder attached to an air pressure sensor. Anyone got any advice for good components to use? I'd need a pressure sensor which works at the appropriate pressure range (i.e. from atmospheric pressure up to atmospheric pressure plus a little bit because the bladder might be inflated plus however much a baby can squeeze), and which has a USB output and comes with USB drivers which provide an API (or some other equally simple way to use the data in my own programs).
To make the air bladder, my best idea would be to use polythene tubing with very thin walls (so it's very flexible). Anyone know a good source for that kind of thing?
What would be nice would be if the pressure sensor was small enough to go inside the tube so that the device trailed only the data wire, rather than a pneumatic tube. We're talking 6 to 12 month olds, so it can't be much more than an inch in diameter.
Many thanks for any suggestions!
Cheers,
Ben
-- Dr. Ben Kenward Department of Psychology Uppsala University, Sweden http://www.benkenward.com
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

I don't know of anything that small...and I understand the desire to buy one ready built
There are many commercially available pressure sensors with analogue outputs, this could be read into the PC via an AD converter and printer or serial port
But the sensor is ~40mm^3, would require a tube/remote mounting
Now, having thunk of it for a few minutes...a hydraulic 'grip thing' might be better than an air 'grip thing' as the pressure from the infants grip (small) would be transmitted more accuratly to a sensor...and the transmitting tube could be quite small, more so to resemble a wire
I'm still thinking of a commercial sensor which has the advantage of being pre calibrated, (SMC, Festo, SunX, etc ) if your own electronics techs can't make you an AD converter this man can advise; http://www.usq.edu.au/faculty/engsurv/staff/StaffDetails.asp?Username=phythian
also http://www.usq.edu.au/faculty/engsurv/staff/StaffDetails.asp?Username=billings may have some unique ideas on the subject.
and a simple program written in some flavour of basic (QB, VB) can read the input and put it in a vector for you to interpret
silicone tubing should be suitable for the grip, it should be readily available from a tubing supplier...i am concerned about losing pressure in tube flex if the total amount to be measured is small, while still providing a 'nice feel'
ANOTHER idea is to use a spring weighted mechanical grip acting on a potentiometer, supply 5V to the pot, read the output and convert...like a teeny tiny version of a body builders gripper
the advantage of this approach is that nothing is lost in turning grip pressure into data, whearas with grip tubes, pressure will be lost in tube expansion in the grip device in order to make it 'feel nice' and maybe in the transmitting tube...both hydraulic and air would be susceptable to leaks
well, thats my $2 worth
Regards,
Reuben Molloy R&D Engineer Control Systems and Machinery Wagners CFT Toowoomba, Qld, Aus, 4350 www.wagner.com.au 0439 206 651
DISCLAIMER Any information contained within this communication, is for the express use of the recipient, and is not to be used in any manner other then the purpose for which it was intended by the sender. Any views expressed or implied in this communication are those of the individual sender, except where the sender specifically states them to be the view of Wagner Investments Pty Ltd and/or its subsidiary companies.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Hi Reuben,
Many thanks for your thoughts. Your reply made me think about a few more issues that I hadn't really considered. In my searches I have come across many pressure sensors, although most of them seem to be unsuitable for one reason or another (mainly price!). It looks like there is probably nothing off the shelf (http://www.midicreator.co.uk/midicreator/sensors.php?id=2 is the closest I found, and it's wrong for several reasons).
I looks like I resign myself a) to building one and b) to that the sensor will be external and there will be a pressure transmitting tube. Given that, the best thing I have found so far is this: http://store.pasco.com/pascostore/showdetl.cfm?&DID=9&Product_IDQ374&Detail=1 . You just have to combine it with a USB converter from the same manufacturer and they are both cheap. I think the problem is that it will only work with their own software (it is for classroom use, after all) rather than them providing a driver with an API.
There is one thing that I didn't mention. I really don't need the thing to be accurate in absolute terms. I just need to be fairly sensitive. I am much more interested in pressure changes than in absolute pressure.
I think hydraulics may be asking for trouble, as this will inevitably be a bodge job (I've built good science on engineering bodge jobs before!). I realised I am also going to need some kind of valve system to keep the air inside topped up. It will need to be only just above 1 bar, or else the poor mites won't be able to squeeze it at all!
Using silicon tubing sounds like a great idea. I can connect it to the pressure tranmitting tube with heat shrink.
I'm not sure the spring pressure sensor will work well because I need it to be able to sense pressure coming from any direction.
So I guess I have narrowed my requirements to: A cheap non-accurate but reasonably sensitive pressure transmitter, either with a USB interface and drivers, or else I would also need a cheap and cheerful USB A to D converter with drivers. Also a good valve for this application, which I can heat shrink to the other end of the squeeze tube.
Any more suggestions very gratefully received!
Cheers,
Ben
-- Dr. Ben Kenward Department of Psychology Uppsala University, Sweden http://www.benkenward.com
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
http://store.pasco.com/pascostore/showdetl.cfm?&DID=9&Product_IDQ374&Detail=1 .
looks like a fine thing...its sensitivity & resolution is good
you could charge your grip with something like a bicycle pump
fish tank plumbing has small on/off valves...pump the grip up, release till desired softness?
heat shrink can be had with built in glue for sealing - regular heat shrink may not seal well enough - consult your heat shrink supplier?
pressure changes would be fine...highest reading - control reading grip pressure
as for how to read it...maybe see your IT guys?
i think USB is serial in a packet...I know nothing of it, but someone around the traps will
API? i suppose so, I am thinking of a program that reads data on the port and outputs it as a vector...start test-read from now until-end test
then you have a box of numbers which will have a lowest value - presumably before any gripping - to highest value, presumably highest pressure attained while gripping
you'd have to watch the test to make sure they didn't roll over and lay on the device or something :)
Regards, Reuben
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Dear Reuben,
Many thanks, so many useful suggestions! Especially the ideas to look for fish tank valves and get special sticky heat shrink.
I going to try this with really young babies who can hardly move themselves yet, so they shouldn't be able to roll over... They might drop the thing though. With older infants cables become a real problem so it's just as well I'm not going to try them.
You've been so helpful that I thought you might be interested in the application: essentially, I want to see if babies can play computer games. That not such a stupid idea as it sounds - there's a debate within infant psychology about whether or not the development of effective motor control is the limiting factor in allowing infants to start to use tools. If it is, then if they somehow had better control over objects at a younger age, then they would be able to perform goal oriented object manipulations - in other words, they might be able to play a really simple game, using the squeeze box as a real time controller. We'll see!
Cheers,
Ben
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Why not go a route different than the pressure sensor, so to speak. There are ribbon type potentiometers that vary thier resistance the more the device is bent. They're very small and a few could easily be stuck inside the toy without much notice at all.
Some links:
http://www.flexpoint.com/products_v2.html http://www.makingthings.com/teleo/cookbook/bendsensor.htm
--Andy P
Ben wrote:

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
scientists!
mad
best of fortune
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

A particularly easy-to-build device:
- a vertical tube or cyclinder, partially filled with tap water, capped at the top, with an input at the bottom.
- a u-shaped resistive wire though the cap, bent at the bottom, and out again through the cap.
- a squeezy bladder full of water feeding into the bottom of the cylinder through a not-too-stretchy tube.
As the bladder is squeezed, the water in the cylinder rises and short-circuits progressivly more of the resistive wire. This lowers the resistance though the ends of the wire, which can be measured with relative ease.
Refinements:
- Since an infant can't be expected to displace much liquid, the cyclinder would normally have to be pretty thin, which would make it a pain to keep the wires apart. Consider pre-attaching the wire to the outside of a solid cylinder that fits inside the first one, keeping the wires apart and reducing the amount of fluid the child must displace.
- If you want this to last, you'll have to choose the wire material and fluid to avoid corrosion.
-- Matt
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Thanks folks for these additional suggestions. I have decided in the end that I'm probably going to use the Phidget pressure sensor:
http://www.phidgets.com/index.php?module=pncommerce&func=itemview&KID 2 9279948129.67.116.6&IID
It works with an interface from the same supplier which goes directly into a USB input, and they provide an API to read the sensors. So I won't have to get my soldering iron out, which is exactly what I was trying to avoid!
Cheers,
Ben
--
Dr. Ben Kenward
Department of Psychology
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Ben wrote:

Phidget's widget?
well the experiment looks easy enough to set up now from the sensor/data end
it'll be like an electronic mobile hanging above the crib!
squeez um gripper and it spins / flashes / makes a noise
i c now...you don't care about measuring the grip pressure, you just want to know the child has gripped
presumably reacting to the stimulus of the game / noise / shiny thing
so when you read the input there will be a floor threshold representing no action, and an 'action' threshold representing 'i have squeezed um thing'
and it will be interesting to keep an eye on how hard they do grip in response to different stimuli
presumably there is an 'good lord, what a scary thing' as well as 'thats so exciting, i want to see it again' type responses
how long do you reckon a baby would last? an hour before it gets tired or bored?
so you might need a good supply of babies!...and of course u'll get to meet all the cute mom's
anyway, have fun.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Ben wrote: <snip> I'm looking for a way to measure

Greetings All (and Ben in particular),
I designed a couple devices along these lines when I worked at NIH a couple years ago. The pneumatic approach is valid, we used it for both human adults and monkeys in MRI environments. We used (pre-amplified) pressure sensors made by Honeywell and by 'All Sensors" Both available from www.digikey.com I liked the differential sensors in the 1 PSI range. Honeywell's ASDX001D44R is one starting place, costd approx $30. One thing I found out is that one can put the sensor *far* from the "squeezee" provided the tube inbetween is of small diameter. We were able to put the sensor 10+ feet away, using 1/16" ID tubing. For sensitivity, you want to maximize (deltaVolume / total air volume), thus tiny diameter where the subject isn't squeezing. In a couple cases, we also filled in part of the "squeezee" to minimize dead air volume, only leaving variable volume where the subject's fingers would actually cause volume change.
One simple way to make the "squeezee" is a short length of vinyl tubing (such as tygon) and plug one side completely, the other plug mated to a smaller diameter tube. (For a quick and dirty version, you can use a hot melt gluegun to plug.) For an infant, I'd try something like a 1/2 -- 3/4" OD tube (thinwall if possible), 2-3" long (roughly the size of an adult's finger -- kids are good at squeezing those!)
A different sensor type to consider are force sensing resistors, such as those made by www.tekscan.com or http://www.interlinkelec.com These would get you around any leakage problems in pneumatics, though they're probably less accurate than temerature-compensated pneumatics. Note that interlink is not (IMHO) very experimenter friendly on the web, and their non-volume products can be hard to find. Here's a link to the spec sheets. http://www.interlinkelec.com/documents/usersguides/fsrguide.pdf (the unit cost is lower than tekscan's I recall. Their FSR kit is IMHO a good starting point: http://interlinkstore.com and then navigate specialty products > OEM Developers Kits >FSR Design Kit ($80 US.)
Both the pneumatic and force-sensing resistors would need some means of analog->digital->your_computer. There are a wide variety of low-cost serial and USB products for such data acquisition. A little googling (I used USB low cost analog OR A2D) turned up lots of hits, one of which is $25 US http://www.dataq.com/news_releases/di148nr.htm I heven't used this so I cannot vouch..... But getting a high-level (couple volt) analog signal in at the rates you'll likely want shouldn't be a big problem.
Good luck with your research,
-- Larry
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Have you looked at wireless truck or car tire pressure gauges? One part goes inside the tire. Possible issues: cost, size, resolution/accuracy.
Mitch
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here. All logos and trade names are the property of their respective owners.