Click to See Complete Forum and Search --> : Does this 802.11b device exist?


afrinl
04-13-2003, 12:28 AM
Hello --

I have been looking for a particular 802.11b device that seems like it ought to exist, yet I can't find one despite extensive searching of the net.

This is the simplest of all possible 802.11b devices, so it seems odd that nobody would have thought of this, so I'm guessing I just haven't found it yet and would appreciate anyone pointing me in the right direction.

The problem I'm trying to solve is tracking patients and mobile equipment in a hospital environment. It may sound silly, but patients often are not where they're expected to be, and a lot of hospital personnel spend a lot of time going to where they expect the patient to be, only to find the patient isn't there, and then they spend more time trying to track down where the patient actually is. It sure would be helpful if we could "press a button" and identify the current vicinity of the patient's location.

There's a similar problem with mobile equipment such as IV pumps, wheelchairs, etc.

Our hospital is about to get outfitted with an 802.11b-compatible wireless infrastructure for purposes of supporting mobile access to our clinical information systems, and it struck me that I ought to be able to use the same infrastructure to solve our patient- and equipment-tracking problems.

Since 802.11b devices are, at their heart, Ethernet devices with their own MAC addresses, you can identify a patient to whom you give an 802.11b device by the device's MAC address. And since you can trace (at least I'm pretty sure you can trace) what wireless access point (which has its own IP address and MAC address) a given wireless client device is using to gain access to the wired network, you therefore can identify the vicinity of any given wireless client device.

So I'm looking for an 802.11b device which would have a tiny footprint (maybe could be packaged in an ID badge you could clip to the patient's gown, or a band you could strap around the patient's wrist, or something like that) and basically would be nothing more than an 802.11b transceiver and a battery (or maybe would work off of a solar cell via ambient light) and would do absolutely nothing more than ping a given server once a minute.

The server therefore ought to be able to know the IP addresses of the devices that are pinging it, and for any given device you want to locate, the server could do a traceroute to the device's IP address, then translate the penultimate IP address in the traceroute (that is, the IP address of the wireless access point nearest the client device at the moment) into the known location of that particular wireless access point.

Ta-da! An 802.11b-based locator device. And given how simple the device is, it oughta be ultra-cheap, too. It wouldn't have to be the most rugged thing in the world, but it would either have to be cheap enough to throw away if it got soiled, or it would have to be autoclavable (i.e., sterilizable). And however it gets attached to the patient, you'd have to be able to detach it fairly easily (with intentions of soon reattaching it) when the patient goes to get an MRI scan.

The same approach ought to work for a mobile equipment locator, except that batteries would become an issue, which is why I raise the possibility of working off of ambient light via a solar cell. Sure, if the piece of equipment (let's hope not the patient!) gets stuffed in a dark closet, the locator device would go dead for lack of power and the server won't be able to contact the equipment, but the server *will* be able to review its logs and see the *last place* the equipment was at before it "disappeared," and the odds would be pretty good that the present location of the equipment is pretty close to the last place registered in the server's logs.

So, has anybody seen or heard of such a patient or equipment tracking device, or does it not exist? Does it not exist because it's so simple and nobody has thought of this application? If so, is there anybody in this forum who's with a company that could build, and would like to build, such a device and use my institution as a testbed?

BTW, I thought of a Bluetooth solution, but our wireless infrastructure probably won't have dense enough coverage (at least, not to begin with) to support reliable Bluetooth access.

Thanks for listening as well as any insights, ideas, advice, etc. you can offer. I would appreciate your copying your response to my e-mail address.

Sincerely,

Larry Afrin, M.D.
Assistant Professor of Medicine, MUSC Division of
Hematology/Oncology
Director of Information Technology, MUSC Hollings Cancer Center
Chairman, MUSC Clinical Information Systems Steering
Committee
Medical University of South Carolina
afrinl@musc.edu

MonkeyFi
04-13-2003, 12:52 PM
Here is a link to something that Symbol makes, it comes close to what you're looking for.

http://www.symbol.com/products/wireless/wireless_sp24_rtl.html

I know I saw a couple of whitepapers on using multiple AP's to calculate probable locations of a transceiver but I'm not having much getting Google to spit them out.

Good luck!

afrinl
04-13-2003, 07:23 PM
Thanks for your note, MonkeyFi. I forgot to mention in my original posting that I had seen the Symbol "RTL" (Real-Time Locator) package, but it's not an acceptable solution for my purposes because it requires deploying a proprietary infrastructure (the last thing I want to do is undertake deployment of a hospital-wide proprietary wireless infrastructure that parallels the 802.11 infrastructure we'll be putting in shortly) and its cost (I found one quote out on the net at a .mil site) is unjustifiable and unscalable for my purposes.

But, again, thanks for your time and effort in thinking about my problem.

-- LBA

dstern
04-14-2003, 03:58 AM
Hello

A software based solution could be build with the Ekahau Positioning Engine ( www.ekahau.com ).
You can use an existing WLAN infrastructure (manufacturer agnostic), but would most likely need to set up some kind of server that receives positioning informations send back from the wireless device on the tracked unit.

David

jatkins679
04-14-2003, 11:47 AM
Originally posted by afrinl
The problem I'm trying to solve is tracking patients and mobile equipment in a hospital environment.

So I'm looking for an 802.11b device which would have a tiny footprint (maybe could be packaged in an ID badge you could clip to the patient's gown, or a band you could strap around the patient's wrist, or something like that)...

I think one thing you may very well have to consider is HIPAA and the overhead required for your IT people. Anytime you begin to talk about sharing patient information (even just names or ID numbers), you have to consider HIPAA requirements. When you begin to talk about sending such data wirelessly, you open a can of worms since physical access to that data stream obviously takes minimal effort.

You mentioned that your system is being setup to access 'clinical information systems', which may or may not mean patient-identifiable and specific information (aka PHI). If so, then it's conceivable that the marginal effort needed for this additional desired functionality might be minimal. If not, then it would be a very big effort to come up with HIPAA-compliant setup from scratch.

In any case, are your IT people up for such a setup? What have they said so far?

yonah
04-14-2003, 02:19 PM
Have you thought about RFID?

I know that there has been a lot of buzz about this, and quite honestly, I don't know much about it, but it is essentially what your are talking about:

Low-power smart tags that can be identified by a myriad of sensors.

afrinl
04-14-2003, 11:44 PM
dstern, jatkins679, yonah --

Thank you for your replies!

Thanks very much, dstern, for the pointer to ekahau.com. Their software solution indeed seems quite nice and appropriate for my problem, and their web site implies they know of the existence of devices which would complement their solution and which are designed for the medical environment, but they don't specifically name any such devices on their site. I have e-mailed them asking for more information.

Thanks for the reminder, jatkins679, but we in fact are already quite aware of HIPAA. We're taking approaches with our soon-to-be-deployed WLAN and our clinical information systems which will be HIPAA-compliant, and certainly we would make sure a solution to our patient-tracking problem would be HIPAA-compliant. However, my fundamental problem remains the same: I can't find an 802.11 device with the characteristics I described in my original post.

That's a good idea about RFID tags, yonah, and we have thought about that, but those systems typically require a sensing network apart from our WLAN, and we really don't want to deploy (let alone maintain) any wireless networks other than our 802.11 network. Also, we're really not interested in being able to sense these tags with "a myriad of sensors" -- I think simple 802.11 communication (device pings a server, server pings a device) is all we need.

Just to keep this thread injected with as much information as possible, I'd like to quote an interesting reply I got privately from Jeff Keenan of Keenan Systems, keenanj@keenansystems.com:

-------------------------------
It sounds like you are looking for a semi passive RFID tag that supports 802.11. These tags will be powered by the signal itself in addition to a
storage battery. It does not look like any are currently on the market but check out this whitepaper. http://www.ing.unipi.it/ew2002/proceedings/001.pdf

These tags would have to support wep, wpa, 802.1x if your network requires them.

You could locate them with a simple SNMP script that polls all access points for associated stations you could also pull signal strength to pinpoint location from the access point.

Jeff Keenan
www.keenansystems.com
-------------------------------

I reviewed the PDF Jeff referenced. It's from an engineer at Symbol, Raj Bridgelall (rbridge@symbol.com), who's obviously been thinking about my problem since at least 2001, yet Symbol's current RTL system really doesn't meet my specifications. I'm going to message Raj and see if he has any news on the particular front I'm looking for since the paper he published last year.

Thanks, again, everybody, for your thoughts on this problem.

-- LBA