[idea] chatbot for scientists doing labwork

(Arne Jenssen) #1

TITLE: Chatbot for laboratory journaling.


Labwork requires a lot of tedious note-taking and journaling. I propose to build a chatbot that can help scientists with the journaling. Then we can leverage the usability of smartphones and build in extra features like image capture.

The chatbot could remind the user at a certain interval to do some measurements and record them.
At the end of the day you can chat with the bot and summarize your labwork.


  • equipment: a sample of a real lab-journal
  • skills needed: programming, experience with labwork, dialog-design


  • Arne Jenssen, full-stack software developer

Please add constructive feedback, suggestions, offers of equipment or skills, and gestures of wild enthusiasm as replies below. Thanks!

(Robert Giessmann) #2

Hi @arne!
I’m just coming home from a day of lectures on bots and data treatment, so it kind of connects.
Giving you some feedback on this from my personal experience: this sounds great for the techies in the lab; this really might help them to get things nicely organized (I have seen people sharing documentation images in WhatsApp groups – it only can get better…) !

Hacking, I guess, is about small iterative steps of improvement, so I encourage you to just start! If you should have time left, though, I would like to direct your attention as well on metadata annotation – maybe you can integrate this into the chatbot as well? – it would help the data scientists a big deal to have nice metadata!

Unfortunately I assume this kind of programming is totally out of my range of knowledge. I can provide a real lab-journal, though :wink:

See you soon!

(Ramin Soleymani) #3

Arne! good to see you hacking again.
I’ve some experience with bot platforms / NLP modules.

I can’t help you full time, but let’s have a chat

1 Like
(Michal Koudelka) #4

Interesting idea. I am not directly involved in lab experiments, but it it’s an interesting topic for me. Let me know if you think you could use a help of another full stack developer.

(Nicolai von Kügelgen) #5

As some writing lab books for a few years now (and also some electronic ones), I think I can give some perspective on that part:

In general this is a really good idea (up to the point that there are several companies/start up’s working on similar projects, e.g. labfolder) and I would really love to see an easily usable open source version for this. There are, however, a few (possibly non obvious) points that complicate usability:

  • Handling: in a lab you can’t just walk around and take notes or pictures with your phone, since you’re wearing gloves and probably handling chemicals with them. When you have to take off the gloves anyway, writing on a computer or actual paper is usually faster and easier than on a handheld device. [This can be worked around by using ‘lab tablets’ directly on the bench. While someone has to pay for them first, they can also directly display experiment protocols and provide direct access to previous notes.]

  • Input encoding & automatisation: An usual labbook not only contains, text, pictures, scribbles, tables and so on - I also uses notifications that may look like gibberish to the untrained eye. Depending on the scientist using it chemical or mathematical formulas, gene names and abbreviations or other technical terms need to be entered easily and without (autocorrect) errors

  • Permanent storage: to satisfy the requirements of both the law and scientific reproducibility a labbook needs to be stored for an very long time, while also preventing tampering of the contents. While backup’s are an relatively easy thing to implement, just leaving out and edit button is neither user friendly nor actually tampering proof. Most electronic labbooks solve this issue by relying on either version control systems (but that requires using plain text or markdown at the back end) or export files with electronic signatures.

1 Like
(Joram) #6

hello fellow labbook expert :slight_smile: I spent a great amount of time introducing people to labfolder in my lab and spent quite some time thinking and talking about electronic lab books.
Maybe we can all have a chat at one point? I’d be very happy to!

(Nicolai von Kügelgen) #7


I’m still waiting for AR technology to get and overlaid lab protocol and pipetting scheme for my experiments ^^