Hi Team,

After a couple months of playing around and testing different setups. I can show you what I’ve managed to put together.

HW wise:

• Raspberry Pi 4
• Raspberry Camera
• USB Mic
• WM8960 Audio Board + speakers
• Fan
• Radar Sensor RCWL 0516 (movement to activate the display)
• AOC 22B2H screen
• MirrorView 4mm Glass from Glas-star.de
• Humidity/Temperature Sensor DHT-11 (to measure room temp/humidity)

Modules:

• MMM-Strava
• MMM-Face-Reco-DNN
• MMM-Spotify
• MMM-PIR-Sensor
• MMM-Carousel
• MMM-COVID19-AMPEL
• MMM-Chart (data from http://opensensemap.org)
• MMM-GoogleAssistant
• MMM-MicVolume (mic deactivated until Face-Reco gets a positive detection, this disables the google assistant - for privacy)
• mmm-systemtemperature
• MMM-WiFiPassword

Extras

• Airplay
• pi-fan-controller

Here are a few pics:

Open issues:
The camera has problems recognising through the glass, the image is too dark. It works… but it requires to a lamp to be on. I haven’t found a way to play with the amount of time it takes for the photo to be taken. The raspistill command does it automatically, but the Face-Reco module doesn’t allow an option for it that I can tell

Thanks @MyMirror

Sure, its a dirty hack as Im not a programmer, but it works.

1. You need to download the data you want from a sensor, to do that, go to the site, select which sensor and data would you like to have and instead of clicking download, right click and copy the link, it would look something like this:
   Options selected: 24hrs data, 10 Minute Value, Arithmetic average, Temperature sensor -> https://api.opensensemap.org//statistics/descriptive?boxid=5b1421ce4cd32e00193f0983&columns=lat,lon,boxName,boxId,unit&download=true&format=csv&from-date=2021-01-03T10:36:32.413Z&operation=arithmeticMean&phenomenon=Temperatur&to-date=2021-01-04T10:36:32.413Z&window=10m

You can later automate this process using wget and changing the date variables in that string

2. The file is a csv file that needs to be changed to json in a format that looks like this:
   [[“DateTime1”, Value1 ], [“DateTime2”, Value2], … ]

There is many python examples on how to do the conversion, but what I found is that they dont return the same format. For example: https://medium.com/@hannah15198/convert-csv-to-json-with-python-b8899c722f6d

The output looks instead something like this:
{
“DateTime1”: “Value1”,
“DateTime2”: “Value2”,
…
}

3. So I take that and apply some command line to replace, move, etc… WIth the above format, you could use this:

grep -v ‘“”’| sed ‘s/"temperatur_20/["20/g’ | sed ‘s/T/ /g’ | sed ‘s/Z"/"/g’ | sed ‘s/",/],/g’ |sed ‘s/": “/”, /g’ | sed ‘s/{/[/g’ | sed ‘s/}/]/g’ |grep 20 | head -c-3

Dont forget to add the first and last [ ]

4. And then you need to enable apache and copy to a location where MMM-Chart can find it

This should do it. I hope it works for you!

Cheers

Hi @emagic,

To be honest, I just followed the instructions here:

https://github.com/paviro/MMM-PIR-Sensor

But something you could try is to turn it on and off in the command line just to make sure that the instruction is correct, as there are 2 possibilities:

/usr/bin/vcgencmd display_power 0
/usr/bin/vcgencmd display_power 1

And
echo ‘standby 0’ | cec-client -s -d 1
echo ‘on 0’ | cec-client -s -d 1

You can find them in the code: https://github.com/paviro/MMM-PIR-Sensor/blob/master/node_helper.js

Depending on that, you need to modify your config.js options

I hope that helps

Hi @drewski3420,

Thanks!
I didnt use any special product, it has a small wooden frame + nails and some aluminium L profiles.
When I ordered the mirror, I had added holes to the design so it would be easier to mount afterwards. You can see it in the pictures here:

How that helps
Stay healthy!
Cheers

Thanks @MyMirror

Sure, its a dirty hack as Im not a programmer, but it works.

1. You need to download the data you want from a sensor, to do that, go to the site, select which sensor and data would you like to have and instead of clicking download, right click and copy the link, it would look something like this:
   Options selected: 24hrs data, 10 Minute Value, Arithmetic average, Temperature sensor -> https://api.opensensemap.org//statistics/descriptive?boxid=5b1421ce4cd32e00193f0983&columns=lat,lon,boxName,boxId,unit&download=true&format=csv&from-date=2021-01-03T10:36:32.413Z&operation=arithmeticMean&phenomenon=Temperatur&to-date=2021-01-04T10:36:32.413Z&window=10m

You can later automate this process using wget and changing the date variables in that string

2. The file is a csv file that needs to be changed to json in a format that looks like this:
   [[“DateTime1”, Value1 ], [“DateTime2”, Value2], … ]

There is many python examples on how to do the conversion, but what I found is that they dont return the same format. For example: https://medium.com/@hannah15198/convert-csv-to-json-with-python-b8899c722f6d

The output looks instead something like this:
{
“DateTime1”: “Value1”,
“DateTime2”: “Value2”,
…
}

3. So I take that and apply some command line to replace, move, etc… WIth the above format, you could use this:

grep -v ‘“”’| sed ‘s/"temperatur_20/["20/g’ | sed ‘s/T/ /g’ | sed ‘s/Z"/"/g’ | sed ‘s/",/],/g’ |sed ‘s/": “/”, /g’ | sed ‘s/{/[/g’ | sed ‘s/}/]/g’ |grep 20 | head -c-3

Dont forget to add the first and last [ ]

4. And then you need to enable apache and copy to a location where MMM-Chart can find it

This should do it. I hope it works for you!

Cheers

Hi Team,

After a couple months of playing around and testing different setups. I can show you what I’ve managed to put together.

HW wise:

• Raspberry Pi 4
• Raspberry Camera
• USB Mic
• WM8960 Audio Board + speakers
• Fan
• Radar Sensor RCWL 0516 (movement to activate the display)
• AOC 22B2H screen
• MirrorView 4mm Glass from Glas-star.de
• Humidity/Temperature Sensor DHT-11 (to measure room temp/humidity)

Modules:

• MMM-Strava
• MMM-Face-Reco-DNN
• MMM-Spotify
• MMM-PIR-Sensor
• MMM-Carousel
• MMM-COVID19-AMPEL
• MMM-Chart (data from http://opensensemap.org)
• MMM-GoogleAssistant
• MMM-MicVolume (mic deactivated until Face-Reco gets a positive detection, this disables the google assistant - for privacy)
• mmm-systemtemperature
• MMM-WiFiPassword

Extras

• Airplay
• pi-fan-controller

Here are a few pics:

Open issues:
The camera has problems recognising through the glass, the image is too dark. It works… but it requires to a lamp to be on. I haven’t found a way to play with the amount of time it takes for the photo to be taken. The raspistill command does it automatically, but the Face-Reco module doesn’t allow an option for it that I can tell

Hi all,

I found myself working on my first Smart Mirror this week and I loved how easy it is to download new modules and have them running in no time.
I was interested in enabling https by creating the certificates and changing the variable under config.js:

useHttps: false -> useHttps: true

Nevertheless I only got a black screen afterwards.

I tried everything I could think of, but, being that Im not much of a programmer, it basically meant hitting the keyboard with one hand while having another coffee hoping that would do it.

Eventually, I managed to find out that I could connect to the https://localhost:port via a browser, which made me realise the keyboard-hitting works… and that it was running… just not with electron.

Well, tonight, after many an hour in front of this, I was able to get it to work, so I can narrow it down to the following:

• Create your certificates correctly
• Make sure you put the right address where they are located
• Not enough coffee
• There is a piece of code in electron.js that seems to be checking the wrong variable:

// if (config[“tls”] !== null && config[“tls”]) {
if (config.useHttps) {
prefix = “https://”;
} else {
prefix = “http://”;
}

The line in “//” was commented out by me. I couldnt find much of a reference to tls anywhere.

• Finally, as the certificate is self signed, these lines had to be added also to electron.js:

app.on(‘certificate-error’, (event, webContents, url, error, certificate,
callback) => {
event.preventDefault();
callback(true);
});

I hope it helps someone else!

Thanks
Regards
JP