Gas cooker on alert system

Hi all, I'm new to the arduino world, I understand some very basic electronics but definately need some direction to start.
My parents, as they get older, on a few ocassions leave the gas hob on, either whilst cooking or just leaving the gas running after cooking. Firstly it's a safety issue and secondly adds to the cost. The UK is insane in energy costs.
I've had a trawl through on the web to see if anything off the shelf is available but haven't found anything myself.
My friend had a few spare ESP01's and also a mega2560, said I could have them to experiment on.
The idea I thought off, which needs some advising on is this.

There are four gas control knobs on the front of the hob. I was thinking of some sensor, like a hall effect could be set underneath all four of them with small magnets attached to the knobs, unless a different sensor is better?
When a knob is turned it, the sensor tells the micro it's on.
The micro starts then counting to lets say 5 mins. The front of the hob has a proximity/pir/ultra sonic sensor and after that 5 mins checks if someone is withing range.
I reckon they need to be within a 200-300mm distance, meaning they are infront of the hob tending to whatever they're cooking.
If the proximity sensor finds no one at 5mins, then a buzzer is emitted, perhaps message sent to their phone sms/app.
Once they realise and tend to the hob, underneath each knob are also push buttons.
Pressing either one will cancel the alarm.
Turning off the knob will of course cancel the procedure for which ever one is being used.

Also, as some cooking times need to be longer and 5 mins could be annoying,
After the first step of turning a knob on, pressing the appropriate push button underneath as before would set the timer to 30mins approx. If the timer lapses, and more cooking time is needed, just need to turn knob off/on and repeat process. I couldn't think of how to extend time any other way.

I could design and 3d print some kind of bus bar that houses the small push switches and sensors. The microcontroller could be trailed to the tiled wall a few feel from the hob. I could then connect it to a steady 5v supply from mains. It could also contain the buzzer/LED etc.

What do you think, does it make sense, feasible? Would the program be easy?
Are the ESP01 or mega2560's usable in this idea?
How can I improve it but also keep it simple in use requiring little or no maintenance.
I'll obviously check and maintain periodically.

Thanks all

Playing with anything gas would require strict certification.
This would be especially true if you introduce electrical controls of any description.
Best leave it alone.

A simple reed switch would be sufficient since you only need to know if they are in the on position or off position

That's perfect. I might be looking for an optical rangefinder (optical ranging sensor) but that might be massively overkill and/or unreliable.

A simpler method would be to simply indicate whether any one of the stove is on or not.
The GE one in my apartment (electric) had an indicator that will light up if any one of the four stoves are turned on.

Yes. Say you are making a pot of soup which might take hours (that does not necessarily need to be attended)
In that case I think there can be a one-time override (e.g. until the next time the stove is approached)

They are perfectly adequate, since you are not using advanced things like USB communication or doing a lot of calculations.

There will be quite a lot of oil buildup from any top surface. You might want to fully shroud it from the top (e.g. leave all the controls pointing downwards)

I don't think he need, as technically he is not making a stove but rather something to add onto it (that does not change any of the gas-related parts)

Yes. You might want to consider user inputs that will not generate any type of spark (e.g. regular pushbuttons). My immediate thoughts include optical sensors and hall effect trackballs.

I would have a MCU watch the stove by using a IR device of some type like a AMG8833 and report on the stove status at regular intervals. The AMG can be feet away and still observe the stove condition. I use 2 AMG's with machine learning, not needed in that case, to identify an animal from a human.

I like the idea of a simple proximity / motion sensor.

Use any means to detect gas is ‘on’, then your idea of 5/10/15 minutes before a buzzer goes off.

Nothing hacked into the gas mechanism.
My concern with IR (it was my initial thought), is that it detects flame / heat… not just if the gas is left on.

I like your idea of hall sensor with a magnet on the knobs.
If any are ‘on’, the timer is started.

You don't need to delete the post. They are perfectly fine.

Open up the window to write a new reply. When you want to insert a quote, stop writing and go select a portion of what any other people have said (drag to select)
Once you do that, a "quote" button will pop up. Click that.
You can also add the quotes manually, in the editing bar. You can insert names by doing

Content that person had said

Yes. You will need to communicate to all of them via like a SPI bus or a I2C bus. Could be quite a hassle.
If you want to do that, you first go and select a desired sensor (MLX90393 for example). Then you look at the datasheet of the sensor and figure out how to communicate to it.
Or, you can look to see if there is a library someone had made (perhaps a very janky library that is horribly structured and missing commands). Depend on the case you might want to do it yourself anyway.

VL6180 "proximity sensing module"
Personally, I will never use something that I can't find a datasheet/schematic for. Unless it's something like a simple transistor or led or the like. You might be comfortable taking whatever code others have for, say, the HC-SR04 ultrasonic distance sensor. That's okay.


I mean, I'm just suggesting. Probably overkill. Might ask your father what's best for them.

This will be Internet Of Things. My current idea would be to have the thing act as a bluetooth device and send a device notice as alarm (or something else). I am not into IoT yet.

Electric won't even have the explosive danger, so it's not a tremendous issue of leaving it on. While leaving the gas stove on is a tremendous issue.

One must assume that any mechanical-electronical mechanism (buttons, toggle switch, potentiometers, etc) will generate sparks. The reason I suggest you use reed switches is because they are all insulated in the glass tubes. You can make crude buttons by 3D printing buttons that have magnetics in them and have reed switches down below. Or capacitive touch, although I haven't seen those used on too much industrial settings.

Take your time. Ask them what would be their preferred method to get warned and take the time researching the sensors and whatever.

The problem with camera is that it's unlikely to be able to see the position of the knob (whether the gas is on or not). It might be able to tell the temperature but it can't differentiate between a cold stove that is not lit and a cold stove that have valves opened.

Wait ...

If your stove is a safety stove (that will shuts off gas when it's not lit), we shouldn't be having this conversation as the stove will shut itself off after it sense that there is no flame. If your stove does not have such safety, get a stove with such safety immediately

If your parents will forget to turn off the stove (have the flame still burning), that's another issue and a thermal (infrared) camera with some very basic object identification can do the job.

if they have a modern gas meter it may have a flashing light to show how much gas is being used and maybe from this you could tell if more gas than expected is being used over time?

to detect if the gas stove is lit a temperature sensor somewhere above the stove compared to one elsewhere in the room should work.

1 Like

The ESP-01 has the capability to connect to WiFi and is massively more capable then the Mega 2560, so it would be perfect for the job.

The Mega 2560 is a cumbersome form factor, expensive, and its only advantage over a UNO is more memory and more pins. Forget it! :roll_eyes:

@bluejets I didn't expect replies so quick, thanks for input.
Yes you're right in regards to electricity and gas, not a good mix at all.
I did think that but as I'm not altering or tapping into the gas controls (not allowed in the UK with only Gas Safe engineers who can check/repair gas appliances).
I'm sure if I use the right sensors and maybe capacitive/resistive buttons, I would eliminate any chance of spark ignition?
With all manner of other appliances/wall sockets within the vacinity, I'd presume they'd would be more of a threat. If I could eliminate the times with gas running (sometimes unlit), I'd be helping.
Appreciate the conerned point.

@cdr_xavier I didn't expect replies so quick, thanks for input. Not sure how to quote your replies in this forum, so will just seperate where appropriate for now, apologies.

1- Ok, I'll look into reed switches instead. Sensors are all new to me.
I would need something robust and reliable, else it defeats the purpose.

2- I'll research into optical ranging sensors too. As you said, perhaps unreliable so lets see what info I find.

3 - The reason I chose the buzzer was due to some hearing loss of my dad plus he'll sit back in the lounge and won't be able to see a warning light, I'm hoping a loud-ish buzzer would work.
Incidentally, somedays he'll leave the refrigerator door open, it also beeps when so, but not loud enough to hear some times (I may need to do something about that too).
Both parents use the phone a lot, are savvy enough using social media so my thinking was a message sent to them or alert (if some pre-made app was available). So, if they're outside the house grounds and have their phone, its another layer of alerting?

4 - yeah, I wish they used electric. I would get an induction hob, saves all the hassle.

5 - Ok, so I can use them micros, excellent. I'll start finding the sensors and will need to learn some code (not my area), though my friend may help and did suggest XOD IDE?

Appreciate the speedy thorough response and suggestions.

6 - You're right about oil build up, even thought controls are not on the top but waist level front facing, they can still get greasy and splashes of liquids etc. I'll need to look into that and placement.

@ Idahowalker Thats an interesting clever idea and probably much simpler. I would need to also check for unlit gas left on, perhaps add a sensor and control both with a micro?

@ lastchancename Thanks for your input. I am a person who thinks logically and likes to plan things through but it's nice to read people understand my process, there I know theres always room for improvement or a smart/efficient way.
Just wish now I was a programmer too, would serve me well now.

Apologies on the layout of my replies. As I've been trying to post the forum algorithms have been popping up on methods to reply also a max of tagging two users as I'm new.

Still not certified and insurance will crucify you, not to mention any building inspections.

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.