Automating/Eliminating Jobs within my Company?

>>60028338
You'll have to wait until you can afford the laboratory information management system. The amount of work that goes into creating something like that is crazy, and if it's even remotely in the minds of your managers, creating a system that competes with an industry standard at a company that's focused on doing something that uses that kind of technology instead of making that kind of technology is a fool's errand. If you succeed, you'll never be able to swap out to the industry standard and now the company has to be a software development company on top of whatever it is you do. If you fail, you've failed and the company has lost money.

>>60028417
Sorry, I should've clarified that I don't want to create something on par with LIMS. It's just that for most, if not all, of the instruments, there seems to be a quick and dirty way to generate reports. I don't want to go the whole nine yards and bring them into the modern era of data management with databases of sample info, but I would like to expedite the reporting process.

One of the instruments outputs the analytical results to an .mdb file, for example. I was thinking instead of having the lab software generate a PDF, which is then printed and then the results input into Excel by hand, there would be some way to extract the relevant info from the .mdb file into Excel (haven't looked into Access or how to use the Excel API yet, though).

If that's not possible, then yeah, I'm shit out of luck.

>>60028338
is it YOUR company? are you getting paid to do it? if it is for your boss, just expect to work on it but eventually kicked out with a pat on the back. they don't care. you shouldn't either.

>>60028480
The last time I checked (long time ago), there were no APIs to properly input data into a microsoft excel file because that wasn't correctly documented by microsoft. So what people did was they use open standards for spreadsheets instead of microsoft ones.

What I would do for automating data stuff (storage and analysis) is writing my own csv files or writing them to a database. Then, I would write my own analysis program instead of learning excel's functions. This allows you to maintain much bigger datasets. Excel was not meant to be used as a database.

If you still need to use microsoft excel, I found this for python: http://stackoverflow.com/questions/23041021/how-to-write-some-data-to-excel-file-xlsx
and this for c#: http://stackoverflow.com/questions/13437727/python-write-to-excel-spreadsheet
and these 2 for nodejs: https://github.com/SheetJS/js-xlsx , https://github.com/guyonroche/exceljs
So I guess I'm was quite outdated.

>>60029190
>continuation 1
My advise for you is to not try and substitute them, I think you may create a bad environment in your company. My advise is to automate all the new things. Let them do what they usually do. Just don't let anyone new come to the company to do more of the same repetitive tasks. Don't let your boss see the need to hire a new guy to enter data in more spreadsheets.

2 types of new things might happen at your company.
1) new tasks need to be done to support new functionalities (new analysis) or new data starts to be collected for the current projects or
2) new projects start, meaning more of the same.
For number 1, automate all this new stuff, let these guys keep doing their tasks manually.
As for number 2, start automating the good old tasks for these new projects only. You will find that, while you develop the data management automation, you can, could or will have to do a couple of things differently. Because no human will be involved in the middle of the process, you will have the opportunity to do things better (imagining these task are very easy, don't forget there are still things only humans can do). These differences create incompatibilities between your's and the old processes. Use them as an excuse to not dismiss these guys, so they don't feel threaten by your work. You don't want them against you, specially if they know how the process work and you don't. If they feel threaten and you ask them any information about how something should be done, be the order something happens or the amount of times one thing should be done, or when, it doesn't matter, they will give you wrong information or won't tell you. At first, you will be seen as doing something unnecessary, so if you don't have spare work time
to do that, you can only hope the work you might have done for case number 1 will enlighten everyone's mind and give you space to automate old tasks.

>>60029269
>continuation 2
How to start? start small, such thing will take months to make sure everything is working fine. Accept if, in the next new project (more of the same tasks), you can't automate everything because you didn't have enough time to automate all the process. Try making things compatible with manual tasks, this is to be easy to go back to manual if needed and also gives you confidence that everything is working, giving the same resuslts. Later on, much further on, you can get rid of the possibility of any step be done manually, substituting these steps for ones where there are clear differences between the manual and the automated.
Just to make things a bit clearer. there's one way a task is done, let's call it way A. Way A is completely manual. In a way B, you can automate that task but just do what a human would do (of course it can't be completely equal but should be equivalent), for the same input, the same output comes out. Writing a program to create excel files is a good example of way B. In a way called C, the process is automated but no human should feel they still can do what your program is doing. Your program has optimized the process to a level where you have to let it do it's work alone forever. The spreadsheet is completely substituted by a database, just as an example.

that's my 2 cents.

>>60029190
>>60029269
>>60029295
Forgot the advise to only live this journey if you really care for that company. In my honest opinion, It's a boss', owner's, parter's job to learn how to enhance their company. Of course they don't know everything about all the things, but not knowing that data management should all be automated, in 2017, is just a case for darwin's natural selection of business. Some survive, some don't.

>deliberately destroying jobs

Grade A autism.