Bottle Filling Level Control System

Is it possible to monitor the filling level of bottles using a vision solution? If so, suggest the appropriate actuator to use. Note that the bottles to be monitored are in motion on a conveyor.

That would require you to provide us with details of the liquid you are filling the bottle with?

1 Like

Can you explain what "vision solution" and "appropriate actuator" supposed to mean?
Are you trying to just monitor or control the filling process?

1 Like

yellow liquid soap

image
Such as the image provided.
By "actuator," I'm referring to, for instance, employing a particular camera.

Sorry, my mistake, I didn't notice this post was on Nicla Vision.
Anyway, consider also some simpler infrared or visible light emitter/receiver setup.
For vision, I would look at OpenCV.

1 Like

Could you elaborate further, please? This is my first time working on an Arduino project, and my knowledge is limited to the basics. Is Nikla Vision a suitable option for this project?

your photo already tells you what to do, but you'd most likely be better off with a raspberrypi + usb camera - detection of the fluid level is quite straightforward. Nicla Vision would work, too, you'll jut need to invest more of your brainpower.

I don't know anything about Nicla Vision, I responded by mistake.
With light beam you could illuminate bottle necks at determined height and detect with sensor on the opposite side. There is difference in absorption between "empty" and "full" and you can measure it.
For this approach, you should open a post in general electronics

You might try doing this with a light beam at some particular height, with multiple sensors on the opposite side. With liquid in the bottle at that height, you'd get a certain kind of spread in the light, basically with the neck of the bottle acting as a lens. If there were no liquid, the readings would be different. Then your program could say "Looks more like this" or "Looks more like that". Of course, a lot depends on the optical characteristics of the liquid and the bottle.

Do you suggest utilizing LDR sensors? That sounds like a good idea, I'm interested. Additionally, I could consider setting up a small dark room to ensure that our measurements aren't affected by daylight or the factory's light.

I tagged Nikla Vision or hazard because it contains the word 'vision.' I've updated the post in the general electronics section. utilizing LDR sensors sounds like a good idea.

There are numerous light sensors, not just LDR. Anyway the key to success is choosing a best wavelenght . Infrared would be good candidate if your bottles are made of pe, since watery liquid is absorbing strongly IR and pe is quite transparent. For pet bottles not so good. Also in visible color the wavelenght counts. One color can be absorbed by the bottle more than an other.
Disadvantage of this approach is that the sensor needs to be quite close to the bottle and protected from other changing light sources

PET bottles I use

NIR around 1000 nm could be good option. For visible light you could do some experimentation with very low power laser pointer, green and red, illuminating through the bottle towards a wall. Which gives more difference between empty and filled neck..?

Apparently you want to move into an industrial system. For filling bottles the wheel has already been invented. You can use a cylinder with a fixed (though adjustable) stroke to pump it in, or suck off the superfluous while filling (and recycle it). Both methods are pretty reliable.

If checking is considered necessary, you already noticed that visual comes with a lot of whats and ifs. Widely used is weighing the bottles.

But he wrote "working on an arduino project". Anyhow, weight scale integration with arduino would be quite easy as well, I made one for legumes dosing.

Just need to control the level of each bottle.

I just tried for curiosity red laser with green dishsoap on transparent bottle.
The laser was little too bright for that but there was noticeable difference between filled/not filled.

1 Like

Could you please test it with edible oil and then share some pictures of the outcome with me?