PDA

View Full Version : How to differentiate between glass and plastic and other materials



karakuli
09-27-2009, 04:33 PM
Hi, I just found this cool forum through google.
I have a school project to design and hopefully build an automatic material sorter. There are 4 materials (aluminum, glass, plastic, steel) and a random object. There are also basic limitations on dimensions

I was thinking about using conductivity to differentiate between steel and aluminum and a light sensor between glass and plastic. The only problem is that I have never used a light sensor before - and is it possible to differentiate between the two? Because the thickness and the color are pretty much random - how can I ensure that 2 don't get mixed up? Would using different wavelengths help?

If you have any other ideas on how to automaticly sort 4 materials, please feel free to share!
Thanks

ScuD
09-28-2009, 11:41 AM
I can't come up with an idea as to differentiating between glass and plastic, but you could use magnetism to decide which is aluminum and which is steel.
An electromagnet would pick up any steel whilst the rest would just pass by, so you could use the transparency to detect which is aluminum and thus have no need for physical touching.

Adrenalynn
09-28-2009, 11:57 AM
Awhile back I posted a cheapie spectral analyzer using an Arduino and some various LEDs. I'll search around and see if I find it.

robologist
09-28-2009, 03:09 PM
Was this it? Or something similar? http://www.creative-technology.net/MAKE.html

Adrenalynn
09-28-2009, 03:20 PM
Similar. In my thread I talked about how he and I corresponded, and the changes I made. Trying to get some xbees to behave, I'll look it up when I have time (or just play with that one. :) )

Obscene
09-28-2009, 09:09 PM
You should build a robot capable of resonating the glass or plastic material and then analyzing data recorded through a microphone to make the differentiation. Perhaps a servo driven finger for tapping and a decent quality percussion microphone?

The main project would be in programming the software. What you could do is, record the finger tapping on various differing plastics and glass and then compile a baseline on the major differences for the software to make comparisons with.

A plastic would naturally make a more "thudly" sound, while the glass would tend to ring out. I think this would be true for almost any shape or thickness. There would be limitations of course.

Adrenalynn
09-28-2009, 10:05 PM
That's an interesting idea. What about hitting it with ultrasound and looking at the returned waveform? Or a really sensitive force sensor on the opposite side?

karakuli
09-29-2009, 08:31 AM
Thanks for the ideas everyone!
I really like a microphone idea, the only problem is that if there is outside noise - could it not interfere with the sound?
Also the bottles can have labels on them, so if a label is right in front of a light sensor - then the whole test will be useless.

I was also thinking about a thermal conductivity test - where I have an insulated chamber, a heating element, and a thermocouple. What do you think? Is it feasible?

Eric
09-29-2009, 08:46 AM
Don't forget the easiest test.... weight. Wouldn't glass be more dense than plastic? It's a goofy appoach to it the problem, but does anyone actually know of transparent plastic weighing more per volume than glass?

Matt
10-02-2009, 08:34 AM
I'll let someone smarter than me ring in on these sensors. If color is a way to do it these might be something to think about.
http://www.taosinc.com/productfamily.aspx?sel=cs

DresnerRobotics
10-02-2009, 09:42 AM
I'll let someone smarter than me ring in on these sensors. If color is a way to do it these might be something to think about.
http://www.taosinc.com/productfamily.aspx?sel=cs

Who are YOU??

ooops
10-02-2009, 12:19 PM
I have a school project to design and hopefully build an automatic material sorter. There are 4 materials (aluminum, glass, plastic, steel) and a random object. There are also basic limitations on dimensions


I know that light sensors and other gadgets are cool, but real world sorters actually "fling" material and distance pretty much narrows down the possibilities of what it is. Glass should fling further than plastic, and steel should "fling" further than alumn.
If you want to keep things high tech looking then fling the material into a pressure sensor and determine the force that it hits the sensor with. Or fling it through a chronograph to determine the speed of the fling.
Of course if you have a big chunk of alum. or plastic, and small chunks of glass or steel you could have problems with the force or speed sensors. I didn't gather from the posts but I am assuming that it would be pretty much alum. and tin cans and plastic and glass bottles?

Good luck with your project, and please keep us posted.

Side-note: for school projects I would always recommend making it noisy (hopefully in a cool way) and incorporating bright lights ... maybe one color light would light up for each of the materials? Get everyones attention when you fire it up with the noise and keep their attention with the bright lights!

mannyr7
10-02-2009, 05:02 PM
Side-note: for school projects I would always recommend making it noisy (hopefully in a cool way) and incorporating bright lights ... maybe one color light would light up for each of the materials? Get everyones attention when you fire it up with the noise and keep their attention with the bright lights!

Great suggestion! I'll have to remember this tip for my kids' future school projects!!

gallamine
01-25-2010, 09:56 AM
Many types of plastic are opaque to infrared light, while most glass is transparent. It will depend on the type of plastic, but if you get several IR LEDs and shine them through the material, you might be able to differentiate the glass from the plastic.