I'm working on an automation project in my lab involving two syringe pumps, four rotary valves, and a chiller. The system needs to handle chemical syntheses and will be running about 70-80% of the time. Most of the time, the equipment will just wait for reactions to happen, but it will run programmed tasks lasting from 2 to 72 hours, like pumping reagents and changing temperatures. I've seen lab equipment controlled by LabView or PLC systems, but not much with Python. Is using Python a reliable option for this? It would save us a lot of money and be easier to program than traditional PLCs. All the equipment has RS232/RS485 ports, and some even have existing Python drivers available on GitHub.
5 Answers
Ultimately, it depends on your risk tolerance. I’ve used Python to control machinery in a steel mill, but we always had a human in the loop just in case something went wrong. It really is about how you design the system and handle failures.
There’s a lot to consider beyond just code reliability here. Think about certifications, quality control, and warranties. This isn’t just a casual project; definitely get some higher-ups involved in the decision-making process.
LOL, trust me on this; we run a synchrotron entirely on Python! Not just for data analysis, but for the whole operation! There are many synchrotrons doing the same. So yes, Python can be super reliable and effective, especially with the right libraries like bluesky that manage data orchestration.
Haha, that's awesome to hear! Thanks for sharing!
This is great news, appreciate it!
Bluesky really does make life easier! I work with it at NSLS-II, and it’s very effective when coded well.
Just to clarify, bluesky is mainly for beamlines, but it's still impressive!
Python can definitely be used to write reliable code, just keep in mind it won’t have the real-time performance of a PLC. If you don’t need microsecond timing, Python is a solid choice and quite efficient for gluing systems together.
The programming language isn't really the issue here. Python can handle plenty of tasks reliably; it all comes down to how well you write your application. Just ensure your code is robust!
Thanks for the heads-up! It’s more of an R&D setup, so no production pressure, and the chemical risks are minimal.