Control your optics bench experiments with Open Source software

[danolo]

Hi,
the aim of this post is to promote the use of Open source software to acquire, visualize, store data of an original optics bench experiment, using hardware accessible from open source drivers.
In general optics bench experiments require software support to:
1) Individually control and acquire data from each mounted devices (motors, camera, lasers, shutters etc..)
2) Automatize your optics bench to perform custom control loops. This involve custom data treatment and heterogeneous device control.

1) is generally easy as when you buy a device from a company, a software is generally provided to control and acquire data from it.
2) is difficult because you do not have necessarily compatible software between the different devices, and you always have to build yourself your own product.

There are commercial frameworks that allow you to do point 2), but you will find after some time that you need a high level of expertise to automatise a new experiment, as it is difficult to re-use existing application developed by others. The bad consequence is that you will become dependent on a particular framework, with rare human expertise spread around ready to help you efficiently. At the end you will face growing time-consuming effort in solving software issues.
So, having made this observation, you can consider the other direction : find and build on top of Open Source software to develop your custom control loops and scans. You will see that this political choice is the best in the long term. Because Open Source software is designed to be compatible with other Open Source software. It is generally designed to be as generic as possible with exhaustive hardware compatibilities. The Open Source community grows constantly bigger and faster, a bit like universe expansion. It is the best sustainable model !
The right behaviour to go in that direction is quite simple:
1) When you buy a device, pay attention to the driver availability. Concretely, prefer those having Linux drivers available.
2) If you foresee getting a student to work in your lab for your new project, he may know already a bit of common scholar languages like c, c++. So orient him on exercising on top of his knowledge and Open Source packages. It will be much more rewarding than discovering a completely new framework that nobody else than him will master after years.
I initiated a web site where I promote useful links to Open Source packages that I used successfully to develop user interfaces to acquire, control and automatise optics bench experiments for various labs: www.opticsbenchui.com
I think it is important that optics people consider strongly which software direction they should go on the long term. They should temper their strong attraction to products promising rapid results. Be careful to what seems to be the easiest path !