Stepwise procedures to guide users through testing

[Sreyanth]

Hi all!

I am Sreyanth, and am a sprint fellow working on developing a system for procedures so that we can post, browse and run our testing procedures in a much better way.

To do so, I started off a couple of weeks ago to understand what different kinds of procedures people do while testing (oil testing, wine testing etc.,) and prepared a list of steps that usually are performed (as a rule of thumb, may be!)

The list of steps I gathered is as follows:

1. Prepare the samples
2. Depending on the material to be tested, decide on the light source (UV, CFL etc.,)
3. Capture a baseline (usually an empty jar)
4. Capture the spectra for the samples (intensities)
5. Apply a filter to smoothen the collected spectra
6. Subtract the baseline for spectral data (absorbance)
7. Select an appropriate measure (like peak height or area under the curve)

Well, what then?

This forms an important source of data for testing. One can compare these with existing spectral database to guess the material (if unknown), or can even know the extent of impurities that are there in the sample.

As usual, I might have missed on a lot of steps, as I based my list on the research notes I read. I would love to hear about the steps I missed on. I guess filtering (smoothing the curves), baseline subtraction and metric calculation can be automated in the above list. What do you think? A part of step 2 can be automated as well? (suggesting the good light sources). 

I request you to send in any ideas you might have to help me out build this system! :)

Sreyanth

​​

--

Sreyantha Chary M

PGP 2014-16 | Member, Consult Club

Dorm 22, Room 26

Indian Institute of Management Ahmedabad, India 380 015

Phone: +91-79-6632-6276, Mobile: +91-812-330-8942

[Jeffrey Warren]

Hi, Sreyanth - one thing that occurs to me is that Step 1 as you've laid out might vary widely -- preparation, i.e. collection, concentration, refinement and so forth could be radically different for different materials or conditions. Are you not yet thinking of that side of a documented method at this time, and want to focus on the software-driven steps first?

Do you have a list of the notes you sourced this list from? I also remembered that the new spec v3 has a great list of some methods people have documented: http://publiclab.org/notes/mathew/01-17-2015/shortlinks-putting-contributors-directly-on-a-kit

Jeff

[Sreyanth]

Hi, Jeff - yes, the step 1 does vary widely depending on the type of testing conditions and materials being tested. Having said that the steps that I gave are abstract steps which are common to all testing methods. 

So when we talk about step 1, it does include collection, concentration, refinement etc., but that is something which I assumed would be simple text, detailing the preparation process. Did I get something wrong here?

I have not taken down all the URLs I referred to, but I did a couple of searches for "testing", "oil testing", "wine testing" etc., and read whatever came up, and took appropriate notes. These URLs include http://publiclab.org/n/10955 and http://publiclab.org/n/10382 which have been documented in spec v3 (thanks for the document! :))

Sreyanth

Thanks

sreyanth

--

Sreyantha Chary M

PGP 2014-16 | Member, Consult Club

Dorm 22, Room 26

Indian Institute of Management Ahmedabad, India 380 015

Phone: +91-79-6632-6276, Mobile: +91-812-330-8942

[Jeffrey Warren]

Yes, I think the non-software steps would be text and pictures. Ok, gotcha, only the ones common to all tests.

Could we present these steps in a table, like a spreadsheet maybe, where each column is a specific test, and each row is a step?