Create access system using HA

Hello, let me tell you about the project that I have in mind so you can give me your comments.

In the building that I live they are considering buying an access system that uses only finger prints and NFC cards.
I found it a little out dated plus I really dislike those fingerprints readers, and I already have built programs that use facial recognition and I’m finishing a licence plate reader, I know different programming languages so I’m thinking that maybe I can build something better.

So first I’m thinking about using what already exist so I wonder if HA with Node Red could be used for something like this. I don’t have much experience using HA but I can learn fast.

Maybe some of you may think that HA is not the best way to go. Or maybe you know a repository that would help for this project.
I also wonder how reliable it would be doing everything wireless using Node red and MQTT. Maybe if I want good reliability I would need to use wires to connect everything.

Please let me know what you think about this project, maybe I’m trying to kill a mosquito with a flamethrower.

This is my idea so far:
Currently they open doors using typical wall switches, so I’m thinking about using sonoff or shellys (but locally connected, I don’t want to rely on internet connection) and I will figure it out how to plug what I have built for facial recognition and licence plate reading so they can turn on and off those switches when a known person or car is recognized.

Thank you

Is this for a multi-party apartment block or something like this ?

Random thoughts:

  • Facial recognition is a very very privacy sensitive topic. A lot of people may (will) object to its use. Especially DIY things where people don’t know where the data is stored and how it’s used. If something doesn’t work or someone who shouldn’t gets in, you will be blamed.
  • License plate recognition cannot be used for access control. Print out someone else’s plate on a piece of cardboard and tape it to the front of your car -> you get in.
  • Legal - if you do this as a DIY project (ie you’re not a licensed professional with insurance) and something goes wrong, you’ll take the heat. Lawsuits, damages, etc. Not fun.

If that’s for your own house, sure, go wild (except for the license plate access, that’s too easy to get around. And ask your family members how they feel about facial recognition before using that). For multi-party buildings, I wouldn’t touch this with a ten feet pole if I were you.

1 Like

I’m shocked that someone would think that biometrics for something like an apartment complex would be a worthwhile investment. Imagine the setup process to capture all those fingerprints. First off biometrics itself isn’t a great idea, its just something you have that’s unique. But its usually not private or has the ability to change, so once compromised the mechanism cannot be used again.

Also what about kids? register your kid for biometric access? Give access to family and friends on a whim too, oh btw you need to submit your prints in case you ever need in…

Anyways principle of something you have and something you know are basic security measures. While biometrics checks off one of those boxes its not easily swapped out or replaced. So a physical key, or rfid access card or token along with a code would be considered for basic security.

Rfid is cheaper than keys, and allows effortless assignment of cards. No wonder why hotels and resorts use assignable cards since they elminate the issues that demand a locksmitch, lost cards and rekeying.

And cameras for access control should be for auditing, not a primary access control. With the current state we are in, getting locked out for wearing a mask or not able to get a haircut would seem foolish.

All this advancement in access control and we still are only protecting against common thieves, any system is easily defauted with a prybar, battering ram, brute force.

Next Topic will be on security systems… :blush:

2 Likes

Thank you both, I wasn’t considering it very well…
Yes this is for an apartment building. Indeed license plate and facial recognition are not good ways of creating an access system. Maybe it’s a cultural thing but no one is worried about the privacy of their fingerprint but I’m sure that it would be different for facial recognition.

So it seems that Rfid and NFC is what I’m going to use. But would that be reliable enough?

I’ve read some tutorials but all of them use external apps that talks with HA. It seems that there is no built in extension for this right? Because I was reading in the docs and it says about NFC and QR tags:

Support for these features is active under development and may not be fully released yet.

What I don’t like about that system is that if a cars with a NFC tag opens the gate, another car could pass and no one would know. So I guess that I’ll have to use AI for car detection to see if anyone pass with no NFC tag.

And yes this protect us from common thieves, but if you really want to break in there are very cheap tools that will unlock almost every door lock in seconds.

A lot of companies use RFID badges for access control and tracking in their facilities. Employees can use their badges to open doors, unlock access to certain floors depending on their access level, etc. The same is true for hotels, which commonly use RFID cards as door keys, elevator access unlocks, etc. So yes, if implemented right, RFID is reliable.

Honestly I would not use HA for something like this. As much as I like HA, it’s very much a tinkerer platform, not something that is anywhere near ready for a mission critical production system. Unless you take very specific precautions (only use core, never update, tightly controlled remote access), HA will often randomly break after an update. There is no stable release channel, it’s like a permanent beta. I would use something with proven long term stability and more mature for a use case where dozens (hundreds ?) of angry tenants will end up on your doorstep if HA crashes and they can’t get into their property anymore.

I would use a commercial product rather than DIY for this, but that’s your call.

This kind of stuff is easy to get up and running for like 95% of the cases. The remaining 5% are very hard to nightmarish. There are a lot of corner cases. Before jumping into a project like that, ask yourself what would happen if it fails and if you can handle the consequences of it going wrong.

In any case, best of luck. Sounds like an interesting project.

1 Like