So it looks the part and sounds the part, but I was just wondering if anyone here actually uses it and if so what for?
I’m going to be looking into data standards and standards in general that are used in the Automation as part of my PhD. Don’t want to be reinventing the wheel trying to get our equipment to talk to each other.
I’ve used it in the past for Tecan Fluents, I was particularly interested in the Tecan provided Python client. Initially I thought it was pretty neat that I could control a Fluent from a Jupyter Notebook but I was also very pleased by the ease of set up. I was able to move from testing communication out of curiosity to implementation of ideas in almost no time.
With that said, it’s fortuitous that you post this today because last week was the initial meeting for a new SiLA Demo Working Group, which aims to give…
scientists a way to test SiLA in their own lab and at the same ease as one can run free trial software could greatly help to establish SiLA as a standard.
This requires a well-chosen collection of pre-configured SiLA Implementations (aka apps) and a demo-workflow wrapped inside a single-file installer.
The demo workflow should be generic and useful even in the absence of special laboratory automation hardware, e.g. guide through protocols, identify and manage samples, read labQR codes, and auto-magically document results.
And so basically the goal is to move from creating standards to implementation of those standards with an emphasis on ease of use, hence a demo. The group is also a good opportunity to scope any further questions you may have. There’s also a Slack channel if you’re interested in that sort of thing.
From this forum, I think that @TimFallonUCSD has explored it as well.
Reviving this thread. 3 years later, is anyone out there using SiLA as the basis for their instrument control? I know UniteLabs is, anyone building their own servers?
I am using it. Here is my collection I use in my current lab: Device Integration · GitLab I know quite some people using it as well, but so far have not met anybody personally using animl.
I went to a SiLA training session ~10 years ago with Travis Lee. David Dabman was there too.
I loved the self-describing, auto-discovery features. But 10 years ago it was a huge pain to spin up a web server and client - esp if you were on a local machine and making a single control application for a device with an existing serial command set. It always seemed like way too much overhead and work.
Nowadays everything is much easier and Claude makes it super easy to spin up a full device server. My team is trying out a bunch of things. I am tentatively hopeful that the SiLA2 interface will be a good abstraction between all the devices and our cloud controller. I still have a few reservations / questions to test out: performance and network overhead, concurrent client connections, how/where to build a protocol execution engine that has good error recovery, etc. We’re actually building out a test integration in our workshop to try it out.
Thanks a lot @Stefan_M for the link. I am hoping to make our repo of servers public too.