RFID expert anyone?

[mikey_p]

Here's the story in short.
Im in my final year at uni and am planning on setting up a very simple RFID network for my final year project.
Context is medical bracelets that some people wear will be replaced by RFID tags and paramedics etc. will be equipped with RFID readers which will be able to read the medicals details of the person either directly from the tag, or, taking an ID number from the tag and retrieving the details from a back end database.
Now here's my problem. I can find tons of information about different tags, frequencies etc. etc. but can find nothing about how to actually implement a small RFID network, I can't even find out what middleware I would need if I was to use a SQL Server DB. I am also looking for info about how to write details to a read/write tag.

I know there are a lot of people on here who work for or own IT firms so if there is anyone that could point me in the right direction for the sort of info I need I would be very grateful. Thanks if you have read this far.

Mike

[dilbert]

The only thing I've read about RFID, in conjunction with SQL, was an article about the propensity for viral propogation. Some American academic did some tests and found that he was able to easily produce a worm that would propogate betwen RFID tags.

Picture the scenario. Everythinbg is hunky dory, your network is running fine, and then someone comes along with a tag that has the worm on it. Suddenly everyone has the worm.

The reason he was suggesting it happened was that SQL would normally be secure behind a conventional firewall, and hence it has never really been considered for the provision of inherent security. Obviously with the RFID tag, depending on the middleware, the database is directly exposed to the real world.

I don't know much about SQL, but his argument sounded plausible to me, but he could be wrong. I wish I could remember what the URL was, then I could post it, but it was a while ago.

[twister]

dilbert said:

I wish I could remember what the URL was, then I could post it, but it was a while ago.

There's this bit of work by some researchers in Amsterdam...

[dilbert]

twister said:

dilbert said:

I wish I could remember what the URL was, then I could post it, but it was a while ago.

There's this bit of work by some researchers in Amsterdam...

Spot on sir.... Hang the American bit, that was the thing I read. It must have been you who posted!

[zumbruk]

twister said:

dilbert said:

I wish I could remember what the URL was, then I could post it, but it was a while ago.

There's this bit of work by some researchers in Amsterdam...

Umm, "this bit of work" is wank.

Please explain how data can contain a virus. Viruses have to be presented to an execution environment. The only way to do that with data is to embed shell-code in the data and overrun the buffer intended to hold that data. An RFID tag only has (up to) 1024 bits of data, and you don't know what architecture the machine reading it has. I don't believe a word of it.

[plasticpig]

You write your own middleware is the simple answer. Generally I would approach this by using a Windows CE powered scanner running over a secure WIFI network or a VPN utilising mobile data cards. Front end on the scanner would be developed in Visual basic. As for the middleware I would be tempted to develop a custom web servie as this could be utilsed in other remote data services. This could be developed in any language but I would use Visual Foxpro because of its inbuilt data handling capabilities.

[mikey_p]

Thanks for all the replys (but keep them coming please, as I research more and more its becoming clear that I may have put myself a little too much into the deep end)
With regard to the middleware, I have found a lot of info on BizTalk Server 2006 release 2 which contains RFID deployment capabilities. If it comes out I think I will be able to get the developer addition free though MSDN, or download the 120 day free trial which should hopefully be enough time to set everything up
Do any of you have experience with BizTalk, is it fairly easy to work with, or would writing my own web service be easier? (but you have to understand I'm not the best of techie's so not so hot on coding)
Currently my thoughts are BizTalk would make it slightly easier to write the event handlers etc. and it provides plug and play for RFID readers etc. However a custom web service may be more suitable as I will only need to take a unique ID, probably put it into a stored procedure or something, retrieve a bit of data, e.g. name, age, medical condition, allergies etc. and send that back to the RFID reader. To begin with, the reader may even be plugged directly into the computer to eliminate some risk.
Also, plasticpig, when you say Windows Ce powered scanner, do you mean a palm pilot for example, with the RFID reader add-on? I have to pester my uni to buy me the equipement and i think they would be more willing to buy a dedicated RFID reader, if this is the case, do you know how the data would be displayed?
I always seem to do this, think of next to near impossible projects. I have no RFID equipement, no experience with it, no software to act as middleware and no experiance in creating web services. Oh well, looks like im going to be busy

Mike

[plasticpig]

I mean something like this bit of kit.

www.symbol.com/mc9000g-rfid

This is the sort of thing wich would actually be used in the situation you talk about. We dont do RFID (in development) at the moment but be we do lots of Barcode 1D/2D stuff and the concept it exaclty the same.

On the cheaper end you would use something like this with a PDA.

www.gaotek.com/index.php?main_page=product_info&cPath=63_96&products_id=328.

Your choice of middleware depends on the scale of your soultion. How terminals are going to be deployed? How many requests are you expecting? If you are deploying across the NHS then you will need some thing like biztalk and a cluster of servers to run it on. If you need it for 25 people then you can look at cheaper alternatives.

[rich-uk]

Microsoft have a few RFID white papers that may be worth looking at...

www.google.co.uk/search?hl=en&safe=off&rls=GGLG%2CGGLG%3A2005-47%2CGGLG%3Aen&q=microsoft+rfid+white+paper&meta=

[fieldl]

dilbert said:

The only thing I've read about RFID, in conjunction with SQL, was an article about the propensity for viral propogation. Some American academic did some tests and found that he was able to easily produce a worm that would propogate betwen RFID tags.

Picture the scenario. Everythinbg is hunky dory, your network is running fine, and then someone comes along with a tag that has the worm on it. Suddenly everyone has the worm.

The reason he was suggesting it happened was that SQL would normally be secure behind a conventional firewall, and hence it has never really been considered for the provision of inherent security. Obviously with the RFID tag, depending on the middleware, the database is directly exposed to the real world.

I don't know much about SQL, but his argument sounded plausible to me, but he could be wrong. I wish I could remember what the URL was, then I could post it, but it was a while ago.

But then we would sell you something like Guardium or Secerno to secure and protect the SQL server.

[vex]

When I did my final year project at Uni we actually got bonus marks for idenifing that the initial solution might be to difficult to impliment and then move over to a simpler solution.

If you feel RFID is going to be difficult to impliment then why not look at bar code scanners. I was fortunate to have lunch with Prof Heinz Wolfe a couple of months ago and he has a trial running using a simple barcode scanner to support everything from medication reminders, on line food ordering and arrival, time/event-logging and departure of support staff.

Having seen it in operation it made incredible sence, effectively removing the need for a mouse or even keypad.

[mikey_p]

vex said:

When I did my final year project at Uni we actually got bonus marks for idenifing that the initial solution might be to difficult to impliment and then move over to a simpler solution.

If you feel RFID is going to be difficult to impliment then why not look at bar code scanners. I was fortunate to have lunch with Prof Heinz Wolfe a couple of months ago and he has a trial running using a simple barcode scanner to support everything from medication reminders, on line food ordering and arrival, time/event-logging and departure of support staff.

Having seen it in operation it made incredible sence, effectively removing the need for a mouse or even keypad.

Yea i've heard that too. I think im going to stick with RFID's as it interests me, and as I want to get into that sort of field when I leave uni it would look good in an interview saying i've set up my own small RFID network. However I have already simplified it a bit, and have a number of idea's which I would like to develop further. For instance I know it won't be as secure as what would be required in the real world, but won't have the time to properly secure it. I have also found a number of points of critisism already for my project. but i'll leave mentioning that to the end.
Thanks for the white papers and other info. Im getting a much better idea of whats needed now.
I think i'll steal a PDA from someone, purchase the RFID SD card add-on jobby and some high-frequency bracelet RFID tags. Create a web service probably in XML, stored procedure and a simple SQL Server DB. Then as if by magic everything will work first time.

Edited by mikey_p on Sunday 19th November 22:08

[spyke]

Have I misunderstood what RFID tags are? I thought when you waved a reader near them, you basically got back a long ID string..? I didn't think anything like an RFID network could exist - that would imply the RFID tags themselves could receive and interpret data.

[plasticpig]

spyke said:

Have I misunderstood what RFID tags are? I thought when you waved a reader near them, you basically got back a long ID string..? I didn't think anything like an RFID network could exist - that would imply the RFID tags themselves could receive and interpret data.

You can have a network if you are using Active RFID as apposed to passive.

[mikey_p]

plasticpig said:

spyke said:

Have I misunderstood what RFID tags are? I thought when you waved a reader near them, you basically got back a long ID string..? I didn't think anything like an RFID network could exist - that would imply the RFID tags themselves could receive and interpret data.

You can have a network if you are using Active RFID as apposed to passive.

Sorry, was using the terminology loosely there. If it was on the large scale real system it would be a network designed to hold the information required and the means of sending it back and forth from database to RFID reader, GPRS would be included in my big design too. What i am creating is a whole lot smaller and won't really be a network, but im just using the term RFID network because it simpler to write.
Also im unsure whether to use active tags or not. There are so many benefits and disadvantages for both passive and active in the context im using. I think passive tags are holding the slightly better hand at the mo though. Due to there longer life, and only holding an ID number which will be useless to anyone who doesn't have access to the database. The active tags would hold all the data about the person meaning it would be potentially more unsecure, but it would be much more reliable as the paramedics wouldn't be relying on a GPRS connection to a central database. Decisions decisions......

[996 turbo]

I'm currently working on this as a sort of side project for my company. There are lots of ways to get a simple RFID project up and running. I came across a company called www.RFsolutions.co.uk who do a simple development kit for about £95. Basically you plug it into a RS-232 enabled PC running some VB software (I think they even give you the source code). This enables you to read or write data to RFID TAGS. You can then use VB / Access to write your data management app. Should be ideal for your uni-project.

However, you probably need to look at ZigBee products (or something like it) to do you Wireless network implementation. These can be used to set-up a local MESH network and actively transfer data point-to-point. You could even have you wrist-band type device measuring vital signs n stuff. It is a completely different route though and lots more work

Steve

[vex]

I can vouch for RF Solutions, I use them and supply dome of thier kit.

[timsta]

Saw this and thought of you....
http://hautegear.com/wearables/intell

[mikey_p]

996 turbo said:

I'm currently working on this as a sort of side project for my company. There are lots of ways to get a simple RFID project up and running. I came across a company called www.RFsolutions.co.uk who do a simple development kit for about £95. Basically you plug it into a RS-232 enabled PC running some VB software (I think they even give you the source code). This enables you to read or write data to RFID TAGS. You can then use VB / Access to write your data management app. Should be ideal for your uni-project.

However, you probably need to look at ZigBee products (or something like it) to do you Wireless network implementation. These can be used to set-up a local MESH network and actively transfer data point-to-point. You could even have you wrist-band type device measuring vital signs n stuff. It is a completely different route though and lots more work

Steve

I can't seem to get onto their site at the moment but I know what you mean. I have a back-up plan of using one that I believe to be similar from Mannings at www.rfidshop.co.uk which only costs £59. If the university is not feeling generous then I was going to purchase this (even with my own money, thats 29 and half pints at the student bar ) and then simulate my idea.
The ZigBee products look interesting. Is that the sort of equipment that is being used to set up these stupidly large wireless networks across cities (i heard Philadelphia is planning on setting up a wireless network across the whole city). If so that could be a good thing to mention for future developments, say for example London implements a city wide wireless network, within London the readers could use a VPN across the network rather then GPRS, which may be quicker? I don't really know what would win between a race to space and back or across a city network being congested by a bunch of kids downloading porn.

[pdV6]

Might be of interest:

www.pcw.co.uk/personal-computer-world/news/2164986/ibracelet-monitors-health-every