Farset Labs Door access


#1

Farset Labs Door entry System

This topic is for the discussion and planning of a community project. Everyone is free to contribute.

Farset Labs is revising their door security system, for entry into the space, and potentially used for internal rooms as well. This topic provides updates on that project, as well as discussion / planning by contributors.


New Door System In Place
#2

Existing topics related to this project:

Pledge towards an arduino based SNARC for door entry
@Jake_Sloan’s 4tronix RFID reader
Original door opening system


#3

FAQ


Can I get involved?

This is a community project, and everyone is free to contribute. There is no one person in charge of this project, so most of the collaboration should be done here.

What is the current plan?

There is no concrete plan yet as to the final design of the system, but there is plenty of hardware available to tinker and experiment with. The access system needs to be reliable, and once complete will not be tampered with except for essential maintenance. The remaining kit will always be available for tinkering / experimenting with improved features.

Currently the plan is to implement a form of RFID entry access for the front door of the space. This is possibly the main priority, but consideration should be given for the use of the system inside the building / on other doors (i.e. no decisions should be made that would cripple it’s use elsewhere in the space)

What was used before?

The previous access system used a web url with an authenticated username and password to open the space. This also allowed for opening the door with an Android app / iPhone Dock shortcut. It’s possible a similar system will be implemented, but there is currently a push towards RFID access, with the purchase of Membership cards.


#4

Repeated on Discourse because @John_Devine has difficulty remembering where I’d asked him to post things after a long long day

Original Post (If anyone doesn’t have access to the closed FB group but wants it, give a director a shout.

The Farset RFID system is up and running and the space is now secure again.

The general member code until cards or whatever wonderful solution our directors greenlight arrives is 5654, this will be the code to enter on the entry pad.

The little blue box above the exit button inside is the main access panel for connecting weird and hopefully non destructive devices to the door system, it contains all the connections from the RFID panel meaning, this would be the point of access NOT THE READER itself as its prone to damage this way. Please don’t remove or tamper with the reader as it took quite alot of work to fix it :S any upcoming projects that need to be connected to the reader can be connected via the blue access box to prevent damage.

I quote a beard “hackability built in is always safer than having someone else take the damned thing apart again.” [Editors note, yup, that was me]

More details and/or schematics to follow soon


#5

It does look damned impressive, and that @John_Devine considered other people might hack it to shit later and made considerations for it might be a Farset Labs first.

Well done guys, this is nuts. Any photos of the current blue box circuitry?


#6

As said on FB, cicuitry and specs/schematics etc coming soon


#7

Kudos to all those involved in getting this running and nice idea using the junction box.

I have been aiming to replicate an updated version of @bolster’s authenticated intertubes entry system (raspi, piface) over the past two days and am mostly there. I would very much appreciate if the wiring diagram/schematic was posted on Discourse as I don’t (won’t) use facebook.

Is the tape on the blue box ‘weatherproofing’ or does it have another function?


#8

Its just there (the blue tape) to discourage hacking it until i get the schematics drawn up. It just means after i get the diagram up, people can hack it with minimal damage / downtime to our security. But it wont be long till then


#9

Hey everyone. As some of you are aware, the current door access system is having a few issues. Getting this fixed asap is key, but we also should be considering our long term solution. After an access issue today I’m aware that the power supply has failed (most likely a wiring problem). The striker has also stiffened and could do with replacement.

This is a rough tasklist of what needs done:

  • Check where the fault is in power wiring
  • Replace/fix power to existing door access system
  • Replace striker with new one if necessary/beneficial
  • Splice striker access into project box to be mounted on wall
  • Work with @martinus/everyone on interfacing the striker from the Pi and future system

Feel free to contribute ideas/suggestions.


#10

This is pretty much the highest current priority and needs to be actually fixed by the weekend at the latest.


#11

Stuck outside again today. Is there anything I can do to help out with getting access fixed?


#12

Photos of the wiring, schematics to come.

[image-gallery]

[/image-gallery]


#13

FYI, there’s a ‘Push to Exit’ button in the top drawer on the wheely drawer unit behind reception. :slight_smile:


#14

Summary: Door fixed - broken relay in keypad.

Issue: The keypad used to enter Farset was damaged. The cause was a faulty mechanical relay on the keypad PCB. The whole system was wired with cat5 and connected using terminal blocks. The following additional issues were identified:

  1. The gauge of the wiring is very small and fragile, it was twisted together rather than soldered. Every connection had a fairly high chance of failure due to oxidation. The terminations of the wires were missing many of the strands due to the insulation being torn off.

  2. Some of the terminal blocks relied on the screw pressing directly onto the conductors rather than the preferred way of using terminal blocks with a built-in plate.

  3. A 6V buzzer was wired into the 12V circuit.

  4. One of the molex connectors on the keypad PCB was damaged/missing. Wires were soldered straight onto the board making removing the keypad for maintenance more difficult than it could have been.

Solution: Thankfully the PCB contained two identical relays so I desoldered both and replaced the hosed one with the working relay. I had to recut and strip back some of the conductors as they had been damaged. All of the wires were tinned before being reconnected and should last longer. Whilst testing was not exhaustive I did test the keypad and manual switch at least 10 times before concluding that it was suitable for use. The schematic below shows the current circuit.

Recommendations:

  1. Replace keypad in the long-term.
  2. Replace the wiring with higher-gauge conductors.

Any questions contact me. :smile:


Any Hackathon events coming up?
#15

Is there a plan to still take this further to rfid level access?

I put together the first rfid reader for SoMakeIt, which just worked on the card’s ID field. It would be spoofable with the right hardware, but worked for us. My plan was to add a pin keypad for a per-user pin to increase security. I think their plan is to go the whole-hog and do a MiFare protocol system with encryption on the cards. This has the advantage of being very secure, but means issuing a new card to each user, whereas using the ID allowed us to use bank cards, work badges, nearly any rfid sticker, etc.

The rfid readers are quite cheap, things like http://www.banggood.com/MIFAIR-RC522-Chip-IC-Card-Induction-Module-RFID-Reader-p-81067.html are <$5 for reading the id.

Mifare readers are a little more expensive, but still frankly cheap - I’d have to check what SoMakeIt were last playing with for Mifare. The software side of mifare is far more complicated due to the encryption.

Anyway, just a thought if more security was required rather than a general pin.

One cool thing about rfid badging in is you can do some cool stuff with integrating “in the space” type of events, though you do have to consider the privacy implications of what is basically entry tracking.


#16

I’ll try and sort this out Friday morning.

My plan of action is simple; get something networked able to unlock the door with some kind of certificate. This will probably be as simple as humanly possible, probably a 20 line flask app which on a /open route initiates the pifacedigitalio module, sets a pin high for a period of time, and responds with a basic set of status responses, as well as a /status route that can be used for any other relevant information including a log field with the past X accesses (if I can be bothered) I’ll probably also add in a /fake_open call that does exactly the same as Open, except it triggers a light or something else so it can be played with easily.

These will be authed with probably just simple basic-auth with a super duper long key, all wrapped up in an SSL cert tunnel. Integrating that with user management is then an open and remotely experiment-able issue. Also means that each ‘calling’ system can be separately logged (i.e. the RFID, Web/mobile based, Internal button somewhere, etc, all talk to this). The door won’t know what user opened it, but it will know what service opened it.

TLDR Low hanging fruit, get door under control, simple api anyone* can use, user access control is a separate issue.


#17

Farset over the past year:

But yes, sounds ideal!


#18

As I said privately in the Digital Infrastructure Hangout, if there is equipment that’s been ‘set aside’ for the door could it be put in my “LIEU” project box so I don’t accidentally grab something that’s earmarked for something else.

@martinus @stealthii no chance you’ll be about tomorrow afternoon or Friday morning?


#19

I’ve prototyped a piface driven door opener based on @bolster’s previous entry system code (thanks to @stealthii for lending me his piface). It’s not terribly involved to implement and could be up and running in a few hours with a few focused people (or one @bolster).

Unfortunately I cannot be there in person to help out with this. I won’t be around for the foreseeable future, sorry.


#20

I’d have thought a piface is overkill (though would certainly suffice). We had a beaglebone so just used that with 12v relay (diode for back-emf clamping and transistor for coil current).

The actual code is a fairly simple NodeJS app that downloads a list of allowed rfid card codes and opens the door if it sees one - https://github.com/somakeit/door-controller