[ESP8285] SE831L intelligent garage door controller - interior, analysis

So, I'm curious about two things. The first - sense. According to the descriptions, it is connected to the gate drive. Generally, probably all drives come with a receiver and remote controls. So what does "smart" mean? Because controlling via the Internet in the case of a garage door seems to me to be more of a disadvantage than an advantage.
The second question - security. In the event of a device break-in, the burglar has access to more than just turning a light bulb on/off. Considering how it works and the price, somehow I don't think safety is a priority here... But maybe I'm wrong?

I don't know if I'm the right person to answer the first question, because I'm interested in these gadgets from a technological point of view, and at home I mainly have "dumb switches" (read: "those that are not intelligent") and the SMART peak at in my workshop is to pair the LED strips above the window and under the cabinets with switch boxes (WiFi really makes it easier here, I use my own SendGet in OBK and Tasmota Device Groups emulation for LED lamps) and adding the option to choose the style of light (warm white / cold white) in home, but:

tmf wrote:

So, I'm curious about two things. The first - sense. According to the descriptions, it is connected to the gate drive. Generally, probably all drives come with a receiver and remote controls. So what does "smart" mean?

Here I would see four potential advantages:
- when we have one application for everything, there is no need to remember about many remotes, even in the case of several family members. Everything is in one place, lighting, cameras, door open sensors, and that titular garage. I have already seen the problem of many remotes at friends and family, then suddenly there is one remote control for LEDs above the "fireplace" (electric with animation), a second remote control for the fireplace, a third remote control for air conditioning, etc... Of course, this also has disadvantages, because what if there is no internet ...
- if someone does not care about privacy, then it can also be connected to the voice assistant and open the gate, e.g. via Alexa
- one ecosystem can have the potential to do some kind of automation like opening the gate turns on the lights or a PIR sensor opening the gate, although many of these examples are not that practical
- I don't know how important it is, but it also gives us a log/log of who opened the gate when, etc., how useful it is depends on the situation and the device itself, there are also ordinary door opening sensors that can also give us such information

However, what it looks like in practice and how often it is used by real users is probably only known by the manufacturer of these gadgets that tracks and collects data, they undoubtedly collect some analytics and probably very rich, knowing life with everything you can, you would have to check what Android Permissions their app gets.

tmf wrote:

The second question - security. In the event of a device break-in, the burglar has access to more than just turning a light bulb on/off. Considering how it works and the price, somehow I don't think safety is a priority here... But maybe I'm wrong?

Absolutely everything that is connected to the Internet is always exposed to access by third parties and should be treated as such. Even those "private" posts or conversations on Facebook, or "private" files on Dropbox or Google Drive may at any time, even as a result of human error, become public for some time.
However, focusing on the device itself from the topic, it would be worth considering first whether we are talking about the version before the firmware change or after the firmware change?

If after changing the firmware, it's probably best to just link this article:
https://tasmota.github.io/docs/Securing-your-IoT-from-hacking/
The above article describes potential attacks and how you can try to protect yourself against them.

If before the change, it's hard to comment on anything else, because the manufacturer's original software is closely related to the cloud and we can never be sure if there is any vulnerability on their servers, at least one unsecured API point, e.g. allowing anyone without logging in to control all devices by their ID or something...

All in all, I can add this, regarding this security, that I have repeatedly downloaded batches from such devices on ESP8266 / ESP8285 and on BK7231 and I saw that both SSID and WiFi network password can be recovered from them.

By the way, the tuya-convert itself for BK7231 works based on... the usual C buffer overflow in the Tuya software and it allows you to upload your own batch to the BK7231 without opening the case (if someone previously obtained function offsets for the firmware build of this device), and this already says something about the Tuya safety standard. If they didn't even secure the buffers in their code, it's scary to think what lurks in the server and cloud software.

Thanks for the info. I actually think the same about it. I am not talking about the general benefits of home automation, but about this particular device. As you wrote, originally it is strongly related to the cloud, so probably whatever it is possible, the producer logs in and does what he wants with it. I think it's worth making the users aware of it. Ideas for including Alexa and similar solutions are IMHO completely off the top. Of course, someone will say - I will be able to give voice commands, but ... we have zero privacy, everything we say goes to the company, and the company does what it wants with it. The question is, do we agree to constant surveillance at the price of unnecessary facilitations? Even if we do, we must be aware that if we do not like the company, our account will be banned and we will not even leave this garage - this is not a hypothetical scenario, because it is already happening. Ew. another version - today it's pretty much free (or rather because we allow ourselves to be eavesdropped and peeked), and when they smell blood, it suddenly turns out that the device requires a subscription to continue working...
IMHO, if you force pack automation everywhere, then only under the condition:
- we change the batch to something we trust somehow (open source),
- absolutely no cloud,
- also works when the net fails.
Since self-replacement of firmware is beyond the capabilities of most users, it is all the more important to make the public aware of the risks associated with such devices, the restriction of freedom and privacy, as well as the introduction of solutions through the back door that normally no one would agree to, e.g. subscription models, in of which we have a physical device, but we do not really own it, because the manufacturer can limit its functionality at any time or simply turn it off.

tmf wrote:

originally it is strongly connected with the cloud, so probably whatever it is possible, the producer logs in and does what he wants with it. I think it's worth making the users aware of it.

After all, the situation is almost identical with other devices, e.g. CCTV IP cameras. Unfortunately, people wanting to make their lives easier forget about their own privacy.

What was that MCU 8051 there for? ESP will easily handle the reed switch and relay.

@tmf I am against all kinds of subscriptions, even those for VOD services, although at the same time I see (even from family members) that some people wouldn't manage without them. A much bigger problem is with subscription software, with software that requires the Internet to work, although it shouldn't need it, with pushing subscriptions wherever possible (where have I seen material about a subscription printer?), and with manufacturers, for example, secretly replacing lifetime subscription licenses:


@krzbor I'm already used to this type of quirks. Below is a photo of an ordinary MiBoxer WL_SW1 smart relay based on the WiFi CB2S module and an external MCU controlling the button and relay, where the MCU communicates with the WiFi module via the TuyaMCU protocol.

Why is there an external MCU connected to the WiFi module via UART (TuyaMCU operates on UART)? Unfortunately, I do not know.

Does the manufacturer warn in the manual about the legal risk of installation? The motorized gate is not a toy and must comply with the requirements of the Machinery Directive. As long as it was controlled locally with a remote control, the operator had supervision over its movement and it did not require additional protection with curtains or pressure sensors. After installing the "over the Internet" control, the status will change.

I attached a small magnet to the gate, and next to the gate I screwed a yunshan plate with a reed switch at the entrance. I have in the app whether the gate is closed. Unfortunately, I often forget "did I close the gate or not?" Sometimes even the remote button is pressed accidentally in a purse or pocket and the gate opens. I fire up the app and see. But that's just reading. I didn't hook up the controls. Still. No one has time to help me, and I'm in a wheelchair and I can't reach the drive under the ceiling. But if I could, I would. Then when I see that the gate is open and no one is home, I could close it remotely. The security problem then unfortunately becomes serious. Though it's nothing to steal from me.

I have 2 garage doors with an electric drive and control with a 433MHz remote control. Half a year I couldn't find time to connect some "Smart" switch to it.
Once I found the time and sat down, I thought calmly, then I concluded that in terms of safety, this is a poor idea.
I am currently using Tuya to control small lighting at home (rather for fun than real functionality) and it happens very often that the relay on/off indicator in the application has the opposite state than in reality.
If I had to install something similar, it would be only with real-time camera preview (although even these can disconnect for no reason ...)

>>20638623
I have the exact (or very similar) model.
And it's been working for a year now.
Because the backyard garage is built of concrete walls about 15 cm thick, unfortunately, it was necessary to add a repeater to the existing WiFi router/modem.
Previously, there were problems with maintaining the connection, now (for about 10 months) nothing like this is happening.
The eWeLink app allows you to open/close/check the status of the gate from any location.
More irritating is the lack of operation of the eWeLink application in my car (with the old Android 4.0.4), because the Chinese probably modernized the soft and at the moment the application prevents logging into the account.
There are no such problems with a smartphone, but you have to use it instead of tapping the big button on the car console screen.

@borba and isn't this Ewelink based on Tuya?

I don't know anything about it.
From what I've read (I'm just an amateur, that's why I haven't uploaded my own software to the WiFi controllers that I have with 9 at home yet), eWelink does not need a central (hub) connecting everything into one.
The principle of operation of these devices is individual connection to the WiFi network, not through a control panel.
So the whole eWelink works for me only synchronized with the home WiFi network.
Regards

Hi I have been a user of a number of Sonoff units around the house for several years. Mainly the Mini R2 but also three of the same as in the subject for the gates in the garages and the entrance gate. I have Sonoff cameras in the garages so I have a view of the garage when opening. I very often open for couriers to leave packages. In addition, there is monitoring of the cameras outside. Since updating the software in these Sonoffs for the gates to the latest 3.6.0, at the entrance gate I have a problem with opening using the remote control. The Sonoff receives a signal from the contracton during opening and unfortunately emits a signal at the output so that the gate stops after opening for several cenimetres. This was not the case before.Only I use sonoff because I have Android Auto. My wife uses the remote control and this is troublesome. Because after opening when it stops you have to trigger it again and stop it so it doesn't close all the way and reach the sensor. After this, the opening goes correctly because the sonoff no longer has a change of state on the contracton and does not issue an output signal unnecessarily. Garage doors do not have this problem....