LECIS - Standard for lab automation

[Markos]

Hi,

It's an old publication, but I believe that the proposal is still valid.

http://jla.sagepub.com/content/4/3/40.full

http://www.lecis.org/about.htm

I think that the proposal has not advanced much in the industrial
environment but could be used by the hacker community which develops
automation projects in the laboratory.

The adoption of this standard could improve the integration of different
lab automation projects using free software and open hardware.

What do you think?

Best regards,
Markos

[John Griessen]

On 02/11/2016 06:40 AM, Markos wrote:
> The adoption of this standard could improve the integration of different lab automation projects using free software and open
> hardware.
>
> What do you think?

Oh... It's had years of development, but it's not aimed at coders very well.

1. It has a concept of state machines built in, where coders want to implement such in their own way.
So adoption by any particular coder, (or hardware designer), requires they learn a new language specifying
state machines when the hardware designers like verilog and VHDL, and the coders like their looping constructs.

2. it was windows all the way: "As part of this project a General-Purpose LECIS Instrument Controller software was written
(running under MS Windows) to serve as an implementation guide",
"goal of this collaborative effort is also to develop a CORBA (Common Request Broker Architecture) and DCOM (Distributed Component
Model) IDL (Interface Definition Language)",
so it was a turn off to open source coders.

CORBA seems to have gone towards CORBA which is open and specs available freely,
but based on object oriented code across networks which is going to be tricky. They even say so: "Why are some specifications so
hard to read?

Most OMG specifications are written for programmers who implement compliant software products. They are not written for the
developers who build applications that use these products, and are especially not written for the users of these applications. "


3., ASTM is one of the old fashioned standards publishers that charge high fees for specs.

My gut feel is CORBA will be hard to really comply with by being so big and backward compatible. It would get you ability to
communicate some with old high priced lab gear, and yet, mostly by serial port in an age where USB or wifi is preferred and even
ethernet is unusual on lab gear, and most of the old gear has no security at all in an age of attackers everywhere on a LAN that
has been infested by trojans from emails.

Just a gut feel though. Prove me wrong!

[Markos]

Hi John,

I do not intend to prove that you are wrong, but rather learn with you. :^)

I'm not fluent in English but I will try to clarify my situation.

Whenever I develop a program for the data acquisition from an instrument
and/or control of an equipment, I have to do some kind of "rework".

I mean, I can take some code snippets, but I can not reuseall program as
part of a system.

It's an old dream to use some method, some standard, some pattern, which
allows me to reuse (and share) my projects as "building blocks" for
future automation projects.

And if I could use, in a simple way, projects developed by others as
part of an automation system.

For example, a few years ago I developed a program for controlling a
valve used in the laboratory. (http://www.c2o.pro.br/automacao/x3192.html)

Later, I developed another project to control a peristaltic pumpin a
bioreactor. (http://www.c2o.pro.br/automacao/x3113.html)

And then another project for data acquisition from an instrument.
(http://www.c2o.pro.br/automacao/x4448.html)

If later I want to develop a system using all these components (valve,
pump and instrument) in an integrated way, I would have to write a new
program, reusing just some code snippets.

The LECIS standardwas more understandable to me, with my level of
knowledge, and allowed me to assimilate some concepts and encouraged me
the idea of doing this integration.

I made already some attempts to use the concepts of LECIS in my
projects, but the challenges seemed too big and I postponed the
challenge and ended up solving the projects in the simplest way, ie
independent program (standalone).

Now, I am encouraged try again to use the concepts of this standard so I
sent this question in this list to discuss it with skilledpeople.

My intention is to use the ideasof this standard, or any other
equivalent, and develop my new project for irrigation with interfaces
which allows to add new modules in the future without much coding
effort, to build more complex systems.

As I understand the Lecis brings a more general specification without
specifying the implementation, while the CORBA concernsto more specific
programming questions.


Imagine if we had an open standard, free, equivalent to LECIS, without
having to pay for ASTM?


What do you think, do you know any open standard that could be a free
alternative to LECIS?


A standard that could be adopted easily by the Makers and help the
integration of their IoT projects?

Many thanks for your comments.

Best regards,
Markos

[J]

Markos <mar...@c2o.pro.br> writes:

> It's an old dream to use some method, some standard, some pattern, which
> allows me to reuse (and share) my projects as "building blocks" for
> future automation projects.

You may be interested by this video on how to make software that you can
reuse:

http://www.pyvideo.org/video/3681/the-well-tempered-api


J.

[John Griessen]

On 02/13/2016 07:37 AM, Markos wrote:
> Imagine if we had an open standard, free, equivalent to LECIS, without having to pay for ASTM?

Sounds good. Simpler would work well for DIY'ers. Then later maybe some AI could learn using it also and skip LECIS?

The python/music/abstraction lecture by Lars Lon: https://www.youtube.com/watch?v=9koJOCL8Bms&feature=youtu.be

is a lesson I could hear.

What steps are repeated most, and how to describe what is in them in terms of old lab gear, USB connected lab gear, ethernet
connected lab gear?

For Lars Lon's style of non-obsolescence, the user needed to know when something was incapable of responding with meaningful data
because he set most server parts of code to "not error out", responding with, "Yeah, sure, I did that", even though they couldn't.
For low cost lab gear, one thing to do differently than his style is security. You probably want a send-message function and a
separate send-message-securely, just so you know it really was secure.

Otherwise, you would like functions abstracted such as "time-difference( rising-edge(machine-22,IC-U2, pin-a12),
(rising-edge(machine-22,IC-U2, a13))", or "measure voltage(machine-22, IC-U2, port-type-adc, pins(a16)).

Then when machines change, you redefine machine-22, which might still use port-type-adc, or maybe port-type-diff-comp instead.
That's the python on microcontrollers version of Lars Lon's abstraction suggestions. It's harder to stay as abstract as he likes
since the hardware is limited and the peripherals and networking are different from case to case.

[John Griessen]

On 02/13/2016 07:37 AM, Markos wrote:

> Imagine if we had an open standard, free, equivalent to LECIS, without having to pay for ASTM?
>
>
> What do you think, do you know any open standard that could be a free alternative to LECIS?
>
>
> A standard that could be adopted easily by the Makers and help the integration of their IoT projects?

I did some searching and cannot find something I heard of awhile back. They were talking about an open standard for materials,
(free published as in no charge to download and read it), similar to how ASTM (which is used by governments relative to building
codes even though citizens can't read it without paying).