Sensor Question

jtrz

Hi all,

Scenario
You have a large pill box, about the size of a small notebook.· It is divided into 28 compartments - 4 rows, 7 compartments in each row.· Each row represents a week of medicine.· Each compartment, a day.· The whole box,·roughly a month.· Each compartment has it's own flip-up door that stays closed with a simple plastic tab and a little bit of pressure.

Goal
We want to be able to sense which compartment's door was opened.· By that I mean "when the user opens the door and points some type of collection device at it" like a RFID reader or something like that.··We do not need a sensor on the hinge or the lock of the door.

Thinking About

• RFID tags in each lid - problem with this include mis-reads due to proximity to the other compartments.· Cannot control what is read.· Or can we?
• Barcode on each door - a possibility with the rigth reader
• IR Emitter - (?) The device with which we want to collect this data has an IR port, similart to those found on smart phones or PDAs for data transfer.· Any way we could use some form of IR emitter and detect that specific compartment based on different IR freqs?

Any other ideas I'm missing?

Any suggestions on how to accomplish this would be greatly appreciated!

Thanks!

John



Post Edited (jtrz) : 4/28/2008 5:49:15 PM GMT

allanlane5

Hmm. You COULD put IR-LED's under each compartement. Then you need a way to 'blink' the IR-LED's in individual patterns.

32 patterns ( 2 * 16) takes 5 bits. Add a 'start' bit and 'stop' bit, that's 7 bits. Say 1 mSec per bit, 10 mSec per 'message', * 28 == 280 mSec. So you only need to point the 'reader' for about a second, and can read about 4 repititions of the message in that time. The "reader" in this case would be a 'wand' with an IR-Detector in it.

Probably the easiest implementation would be using a '164 serial to parallel output chips, along with Darlington Arrays, to select the desired IR-LED. Then run a common I/O pin to all 28 IR-LED's. Then do a "FREQOUT IR_LED, 1, 38500" on the one pin, repeated as needed for the bit pattern. The idea is that ONLY the 'enabled' IR-LED would actually flash, all the others would remain disabled. Then select the NEXT LED, and give its pattern.

DosManDan

1)Just to clarify, you want to know when the door is opened, not when the medication is removed?
2)What happens if the door is opened and nothing is removed? For instance, is this a security device so you can determine if tampering occured?
3)Is the box made from clear material, like alot of pill containers?
4)How small are the compartments? (height, width, depth)
5)Is the box anchored down or is it mobile?

Thanks,
Dan

allanlane5

They do make a 'copper tape', with adhesive on one side.

So, you could put a strip of copper on the outside, and another strip on the inside. The person, when they open the lid, briefly squeezes the two strips of copper toward each other. This 'connects' the two strips with about 10 megohms of resistance. Or they could use a pair of tweezers to grip the lid, thus 'shorting' the two strips together.

You can wire all the 'outside' strips together to a BS2 I/O pin, then wire the 'inside' strips to a set of '164 serial to parallel chips. Scan one bit through the array until you see the 'blip' on the input pin.

jtrz

Hi Dan,

Here you go...

1)Just to clarify, you want to know when the door is opened, not when the medication is removed?
No.· Even easier than that.· I want to be able to have the user scan the door that IS open.· I don't care about the door being opened or when the med is removed.· For example...our one idea was to have a barcode under the lid of the door.· When the door is opened, the user scans the barcode and that tells us what door was opened.· Not too concerned about when med was removed or exactly when the door was opened.· Like I sad, nothing needed in the case of the hinge or the clasp of the door.·

2)What happens if the door is opened and nothing is removed? For instance, is this a security device so you can determine if tampering occured?
Nothing.· We don't care about that.· We're more concerned with knowing the compartment the user is scanning.

3)Is the box made from clear material, like alot of pill containers?
Yes.· Looks a little opaque.· But essentially clear.

4)How small are the compartments? (height, width, depth)
Max?· from 1" x 1" x 1"· to about· 1.5" x 1.5" x 1.5"

5)Is the box anchored down or is it mobile?
Mobile.· Right now it's just a plastic box with NO electronics in or around it.

Hope that clarifies things!

Thanks,

John

DosManDan

I think you have probably stated the answer already. I would use a small 3 of 9 barcode on the inside of the door. Then use a wand barcode scanner.

This one is about $71.00 and I've used these before. They work well in small spaces where you might have several codes close to each other.

https://www.binaryplaza.com/products.aspx/American-Microsystems---Opticon-Wand-barcode-scanner-WND-OPV910-Input-Devices-Bar-Code-Readers~~MPN=WND-OPV910~~M002987571

Kind of curious, why do you just need to know the door was opened? I'm wondering if there is a better way of looking at what you are doing. It is odd that you don't want to know if anything is removed, so I figure you must be trying to solve some inventory issue with a manual count.

Best of luck with your project, and let us know what you decide to do.
Dan

John Burrow

John,

This may be a little off topic, but I assume this will· Not· be for prescription medication. Federal law requires that prescription medication is kept in the original container, with original label. Not that anyone will know, unless to you try to get on a plane.

A little more on topic - I have some experience of RFID, and I think that with tags in close proximity you will get unreliable results. I'd try barcoding.

Good luck, John

jtrz

Hi Dan
Thanks for the reply. Not dealing with pill counts. And actually, we don't really need to know the door was "physically" opened or we'd have a sensor on the door/hinge/clasp. This is for medication "compliance." The patient opens the door, wands the barcode connected to our small hand-held computer, then takes their pills. To the best of our ability, we now we know the compartment they took them out of and when they did. We can also track the day they are on in their regimen. If we have a reference to the medication IN that compartment, we can also track which meds they took that day.

Dan (or anyone) - Do you know of a small, pen-like barcode reader that does NOT require the user to swipe it across the barcode? I'm thinking of something like the pen reference you gave me but acts as a barcode "gun" OR could read 2D barcode labels. Just want to eliminate the movement of the wand if necessary for the patient and make it a point-and-click thing for them. Otherwise, this is a good cheap solution and I thank you for the reference.

Hi John
Thank you too for the reply. When you are working with patients that, for example, are transplant patients that have medication regimens that include like 15-25 daily drugs, most of which are for anti-rejection purposes, the standard way they organize their meds is to have a pill organizer set up for them like the one I describe above in the original post. They have to. If they tried to do it at pill-time, forget it. Their regimens are generally too complex to remember. There are a ton of problems with that. So they (or their caretakers, or the hospital staff) organize ahead of time using that type of pill organizer. Not much different from the ones our parents/grand-parents might use on a weekly basis, only larger.

BTW: I agree with you on the proximity of the RFID tags.

Amaral

Just wondering, If the user opens the door, take the barcode reader , press it in front of the inside part of the door... This is not really mobile technology !

Why isn't a push botton inside the compartment , the guy opens it , press the button, take or leave the pills. You just need to scan the push buttons, just like a keyboard does !

I'm I completely out here ? or that makes sense ?

Just trying to help

Amaral

jtrz

Thanks Amaral. The way you are thinking about it makes sense. But you are going to an extreme with the definition of "mobile" for this solution.

It depends on how you define "mobile." But actually, yes it is mobile tech. It depends on (a) the situation for the user (do they need to be able to roam around with this on their person, in their car, etc) and (b) what type of computer is this system running on (a stationary desktop PC in the home, a laptop, or something like a Smartphone? In our case, it's more like how do we bring this technology TO the user and WHERE (home ideally, hospital, clinic, etc).

We're not looking for this to be a totally self-contained custom device. We are trying to keep the costs low by piecing things together to meet our goals.

As far as the push buttons go, I like your idea. Never thought of that. Thanks for the idea!

DosManDan

One reason I'm trying to help is that I deal with hospital systems that do pill dispensing. Now that I know what you are trying to do, why not solve this the way we do. We have a machine that is loaded with pills. Each pill type goes into a dispenser container. All of the containers go into a machine.

You could use a rotary dispenser with a stepper motor to properly position the cylinder. Then, open the door and the pill falls out. If you create a container with depressions for each pill, then load the cylinder into a tube, you would only need to have one door.

The Micon would keep the inventory, and of course when the pill was dispensed. The machine we use prints a label, heat seals a bag with the meds and dispenses them for use on the ward.

Hope this helps,
Dan

jtrz

Hi Dan

Having a problem understanding your description. Sounds interesting. I just cannot picture it probably because I'm a software developer and not a pharma/healthcare person. Can you contact me at jtrz at com{remove this}cast dot net so we could possibly discuss more off-forum?

The reason I'm on this other solution is that the 28-day pillbox is what was presented to us for a solution. We weren't trying to solve it from the ground up like that. We have another, more flexible way of handling dispensing medication and tracking compliance that is not "container-centric" as this is suggests. We were contacted with this container because of our existing compliance tracking system.

Thanks,

John

Post Edited (jtrz) : 5/1/2008 3:54:28 AM GMT

DosManDan

I sent you an e-mail from my other account.

I look forward to hearing from you,
Dan