Can I plug an arduino into the usb port of an offline computer, and automatically share files moved to it to a network folder/google drive over wifi?

This may be a naive question, I've never tried this before.

I work in a lab where there are 20+ PCR machines each connected to a local, off-network PC hard drive & monitor which run software for the machine, and to collect & analyze data etc. The machines are hooked up to the computers via ethernet and the computer treats it like a network (no internet)

The problem is these computers/the software has totally locked out all internet functionality. Ive figured out that the software basically bricks the connectivity features at the OS level and has accessory software available (that we cannot have) that handles all the networking features itself to ensure security (eg, they created a problem on purpose and want to be paid to fix it).

Users are going up and down stairs through multiple airlocks repetitively all day to save their individual .txt data files onto a flash drive, carrying it upstairs to networked computers and moving them to a shared server. We frequently have failures of communication leading to different people collecting the same data multiple times and we generate hundreds of files a day. It is becoming ridiculous.

What I am hoping to achieve is to have a raspberry pi work virtually exactly like a flash drive does; plug 1 end of a USB cable into the device and the other into the computers usb port, have the software offer it as a location to save on like normal etc.

Then I want the device to automatically upload any new data to a shared network folder or a google drive over wifi and delete files from itself once they're on the network.

Ive also considered getting a wifi-capable zip drive for every computer and just having them constantly plugged in, and connecting all of them to an arduino device capable of producing its own network independent from the work network. Then the arduino could copy any new files it finds to a network accessible location.

Is this a thing that could work as described? If so I would love some advice on what hardware to go for and/or any resources that might help. I've found a lot already but none of it describes exactly the thing I'm hoping to make. Thank you!

Usually the things you describe have been set up that way for a purpose. If you are the director of the lab, why not have it changed the way you want?

I've been told to find a way to get around it that doesn't involve making changes to validated software.

It sounds like you are saying:

  • you have 20+ PCR instruments.
  • Each one is connected to its own PC with HD and monitor. So there are 20+ PC computers.
  • You can copy files from each PC to a USB drive using ... File manager?

Are all 20+ instruments all at the same location (room, floor etc)?

Do you think you could copy files to a USB Serial port? Perhaps using a program like Putty.

This will not get you all the way but its a place to start seeing what might be possible.

Lol yes 20+ computers, and no unfortunately they arent in the same room, they're in 4 different rooms each seperated by no fewer than 3 very slow airlock doors that only allow one to be open at a time, one hallway has an intersection of 4 doors in an extremely high traffic area. Each room has a different gowning requirement to keep contamination downstream of the experiment setup area, so it is EXTREMELY time consuming to keep physically retrieving the files. Plus, if someone retrieves a particular data file but doesnt inform anyone, we frequently end up with multiple operators retrieving and performing analysis in duplicate, which wastes hours every day. If the files just dumped into a common network folder automatically, we could all just instantly know whats been exported and whats already being worked on.

I have driven myself nuts trying to find the directory it saves experiment/results files to, but the root file for user-generated data is not visible in the file explorer. We save to USBs from the software when it prompts for an export file path, then gown back through a cascade of airlocks, up the stairs and immediately move the files to the network or start analyzing it locally.

Here is the most comprehensive manual I found for the most frequently used machine. The other machines aren't substantially different within the scope of my problem.

LC 480 Operator Manual

If you search for the word 'network' you will find the disclaimer warning stating that the instrument must not be connected to a network, although network hardware is present.

The reason they have written this & configured the instrument software to enforce it is to keep the machine compatible with labs who are legally obligated to comply with the strictest FDA regulations & labs subject to privacy obligations related to medical information.

My lab is research use only so we do NOT need to operate off-network, but we also can't upgrade to the network-configured software without going through a ton of red tape and expense that we simply dont have the bandwith for.

In chapter 12 it discusses installing the software in local only configuration, and says to contact the manufacturer for details on network capability. When you do this they try to sell you their LIMS system which essentially achieves network functionality in such a well-controlled way that it still complies with the highest level of regulatory scrutiny, and is expensive and not practical, especially if I can find a diy solution.

If anybody is interested in the labyrinthine regulatory landscape for genotyping kits, I have enough links bookmarked to make you wish you'd never asked. Tl;dr a lot of the usual rules don't apply in our case.

Do you have WiFi reachable from the stations?

There are commercial products: