Who does the RFID door & other machines access? Knowledge transfer request.

[Patrick]

Hi,

   I'm planing to use RFID based authorisation system myself & I'd like to get in touch with the monkeys that did / maintain the system for LHS. In particular I'm interested in how the data for access authorisation is being send to the main access server. I hate an idea that at each reader I'd have to put a Raspberry Pi or anything so ridiculously powerful just to open the door. I'm interested in protocols used to transfer reader's data and commands to make an action. Readers, I have, use Wiegand 26/34 bit standard format for data but do I transport that over RS-485 / RS-422 / ... (?) or TLS'd connection over LAN is the best solution? What are the problems that you had to overcome and I should to keep in mind?

   For some reason I like to code in Assembly / C on Microchip controllers and I'd like to use my experience to do this project. I am not interested in Arduinos or any other preassembled kits capable of GigaFlops and so happening to have an Ethernet connector. It's down to the custom metal and Assembly time running on couple of kHz :)

Patrick

[Tim Reynolds]

What you're describing is basically the opposite of our door system, sadly.

We have either full PCs (!) or RPIs at each door, with a USB RFID reader
and an arduino attached. The machine downloads a list of active cards,
checks each card presented to the reader and uses the arduino to trigger
the door solenoid. The arduino also handles the doorbell and
denied/granted LEDs if fitted.

This wiki page probably explains it better:

https://wiki.london.hackspace.org.uk/view/Door_control_system

> --
> You received this message because you are subscribed to the Google
> Groups "London Hackspace" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to london-hack-sp...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.

[Kimball Johnson]

Hi Patrick,

The doors we have at HacMan are controlled with a Weigand protocol reader, so we can probably help a little.  Take a look at https://github.com/HACManchester/ALF

Kimball

--

[mikethebee]

At Reading RLab we used Arduino based readers on the door. Several systems were evaluated before the boffins chose that method, so I guess it is the most reliable and cost effective way. Having a microcontroller on each door allows for local caching of card info, and this speeds up the response time for recently used access codes and gives some protection against the loss or delayed server access. If you want more details I suggest asking on https://groups.google.com/forum/#!forum/reading-hackspace -Mike

[Mark Steward]

The last thing you want with an access control system is to design it around one hero who is the only person able to maintain it (that's separate to only having one person making changes). Our decisions have prioritised:

 - how cheap and off-the-shelf it is

 - how easily parts can be swapped out

 - how easy it is to debug over the internet or by instructing someone who's never touched it before

Raspberry Pi's (or similar) are the easiest way to get an SSL-speaking device onto the network. You can make an Arduino speak Ethernet, but it's slow and there's not a massive community of people working out bugs. The Pi's log over the network to avoid writing to the SD card, as power cuts can result in incomplete writes. However, as we're using Linux, we could just swap in an old PC if we were having real trouble.

We use USB HF RFID readers because modules at the time were £50+, although there are now nice modules for a few pounds. We used an Arduino for a long time for the doorbell and a simple Darlington circuit to fire the solenoid, but the Arduino has finally died so we're now using USB relays off Ebay. If you want to roll your own electronics, this is where I'd recommend spending your time - but document it and make duplicates of everything!

I've updated the wiki for reference.

Mark

--

[Patrick]

Wow,

   Thank you gentleman. You have deferentially raised couple of issues I didn't have on my "to consider" sheet. We all know what a pain the lack of standards for automation is. Security vs usability vs reliability vs cost ...

   Thank you. I have some thinking to do about the design.

Patrick