New Scio detector might allow automatic detection of materials


#1

The newly crowdfunded Scio near infrared detector might be just what is needed to fingerprint new materials automatically. https://www.consumerphysics.com/myscio/ These detectors are quite good at non-invasively scanning a material and identifying them. It so happens that it works best for materials Glowforge can utilize.


Material identification
#2

Wow, that’s amazing!!


#3

I think it could be a game changing technology for GF. It could be incorporated as an add on or hopefully a basic feature.


#4

Oh man, having it as an add-on would be so useful- imagine all the random scraps you could test before even hitting them with the laser!


#5

With a small gas sampling device, it might give early warning on hazardous off gassing products of a new material. I think Glowforge could blow away the competition once again.


#6

I’m pretty skeptical of the claims that SCiO is making there… I don’t doubt that their device can function on some level, but replacing cubic-meter lab equipment with a cubic-centimeter sensor doesn’t score high on my feasibility meter.


#7

I can haz tri-corder?


#8

Until you start noticing that all your material seems to be the same, because all the GF material comes with a wrap to protect it from surface burning.


#9

I’m going to be skeptically optimistic about this product. There claims are pretty fantastic, for a hand held device. Until I see some testing data that shows its accuracy and reliability, I’ll stick with cutting what I know.


#10

A five megabyte hard drive used to take a truck and a forklift to move, but…
The lab units they compare to are gas spectrometers where a sample is vaporized and its spectra anaylized - so I seriously doubt the accuracy is comparable.

Never the less, very cool to deliver that capability to the public!


#11

Correct, a surface detector can only scan the surface.
Willing to bet I could “fake it out”.


#12

There are some interesting specs on the FAQ:

I’m interested in this as a solution for another project I am working on.

How accurate are SCiO’s analysis results?
SCiO typically detects materials in concentrations of 1% or higher. Concentration levels of 0.1% or less may also be feasible for some materials, however this is rare. The exact specifications depend on the application and material being analyzed.

At what distance should SCiO be used to scan an object?
SCiO should be placed up to about 20 mm (~0.8") from the sample. The recommended distance is around 5mm (0.2"). Objects should never be touched directly during a scan.


#13

I agree something doesn’t sound right about this. It reminds me of the sport band from a few years ago that claimed to be able to track your food intake from scanning your skin. That never made it past the the design phase since they never got it to work like they claimed.


#14

These are already available for purchase.


#15

Well then, buy one, test it out and get back to us on how it works.


#16

They are available for preorder, but who knows if the final product even exists yet.

I preordered the vessyl drinking glass that is supposed to track what you drink a few years ago and there still is no product yet, sometimes companies convey their idea more then the reality during preorders


#17

They only mention having a spectrographic camera in the info-mercial linked.

So… it will depend on the lighting in which you are making the measurement. It will depend on the other colors in frame. Ideally the distance cuts that out as a relevant concern, but the color of the object itself can cause some minor issues. Though often the color is relevant information to the material, so not a total loss there. The lack of light source built in to the device is the reason they say you cannot place it in direct contact by the way. No light bouncing off means no signal to analyze.

Really it will be almost entirely a color detector now that I think about how it works. That is precisely why most spectrometers vaporize the sample in question, then shine a true full spectrum light, or just turn the vaporized cloud into plasma so it emits light itself.

The database is the main thing that has me raising an eyebrow. Even with the surface limitations, there is a fair amount you can tell. There are myriad variations on “green” which can each tell you something about what is in the object. And you can do a scratch test to get the sub-surface components exposed. But with too large of a database, you will have issues with filtering. Some will be easy – “You either just scanned the hair of a dalmation, or you scanned plywood coated with prozac”. But other ones may not be so easy – “This medicine contains Acetaminophen, or Cyanide”

(Do note, in each case I am making up random things, I did not look up spectrographic data on the items listed to ensure they would overlap)

The point is… a spectrograph tells you “There was a peak in this specific set of wavelengths of light, that corresponds to these chemical elements” And biological components are not made of highly varied chemical components.


#18

I plan to. I have an air sensitive compound that I need to monitor during production. The Scio can see and analyze through this glass flask with its near IR illumination.


#19

@jacobturner The technology section states that it contains a near IR illumination source.


#20

They exist to the same extent as glowforge does. They have various prototypes that have been reviewed online and are in first round beta testing.