Deciding on a Knowledge Archive System 1

[code1]

I am in the process of setting up a knowledge archive system on my company servers.

Among material I want to storing/ archiving: Engineering calculation sheets, drawings, case studies, results of plant comissioning and monitoring, vendor data, company and customer specifications, Codes and Standards, design information, etc... Basically anything goes, if it is relavant to what my company does.

It will be like the window's Explorer Tree system for files. Suggestions please what to include at Titles of "Main" folders and their "sub-folders". How expansive and encompassing to be?

Thanks in advance

 

[MintJulep]

No file system will do what is really needed.

To be successful you need to develop a system to accurately categorize each document. Categories should include things like: project #, client name, type of calculation, key words, part numbers and names, supplier names, document originator, and on and on.

This all needs to go into a database to allow you to search for combinations of categories to find what you are looking for.

Years from now the question will be "Remember that we used that gizmo from Finland once? Do you think that would work here?"

 

[kchayfie]

If you would like to learn from the mistake made at our company...

Make sure that the archive is populated by someone with experience and understanding of your documentation and the patience to fill out things like "key words" and accurate document descriptions.

When my company decided to archive that sort of information electronically, it seemed like a good idea to hire a couple of tech clerks to go through the dull and simple task of scanning in the paper archive. Because of a) the type of person we ended up with and b) the fact that they had no idea what they were looking at, we ended up with a database of unidentifiable and unlocatable documents that no one uses and was a big old waste of money.

We've now got one of our retired senior engineers back in for a few mornings a week (to avoid the wife!) to go through and see if he can file things under useful headings and with searchable keywords so that the grand idea of "knowledge management and sharing" can be resurrected again. We've yet to see if this is going to be used...

 

[zdas04]

Kchayfie has hit the essence of the problem.

For another example, I designed a database in the early '80s to catalog the 500,000 documents that were required to be instantly retreivable for a multi-billion dollar lawsuit (that eventually went to the Supreme Court) before the days of PC's or even high-volume scanners. The database was elegant, with nearly infinte flexibility and really extensive (for 1982) search capability, then they hired a platoon of $6/hour paralegals to try to summarize and keyword these highly technical documents and correspondence. About a month before the trial was to start, the lead attorney tried to do a dry run with documents that he knew were there and nothing came up. The emergency kicked off and hundreds of company engineers spending the next 3 weeks sitting at mainframe terminals trying to fix the data. Many millions of dollars later, the document-retreival system (several hotel rooms across from the courthouse full of file cabinets) ended up being a key part of winning the suit.

Bottom line is that capturing and organizing the data is reasonably easy, but without a high-quality data collection and management plan it will fail.

David

 

[IRstuff]

The previous posts are a start, but even with decent catalogging, slogging through reams of virutal paper is still pretty daunting when you want to find a particular answer to a particular problem. I've got thousands of articles on different subjects, but 99% of them are peripheral or even irrelevant most of the time.

I would suggest that you have an internal Wikipedia, perhaps also populated via those retired engineers. Articles of general interest would be created initially, followed by more specific types of articles.


TTFN

FAQ731-376

 

[KENAT]

We have an internal Wiki, it hasn't caught on. In fact fact pretty much the only posts in the Mechanical area are either about CAD or a few company standards that I've helped introduce. Probably 3/4 of the posts are either mine or from interns I've asked to post something.

 

[swertel]

We tried an online forum-type knowledge base. Basically, the only people who post are those who moderate. No one else posts. They went there initially expecting it to be populated with all the answers they were looking for. When the answers weren't there, they never went back.

"It's not my job to add data to the forums."
"How am I supposed to know what's valuable to add?"
"If this system was that important, management would sent out an email telling everyone what to add."

And the list of excuses goes on.

--Scott

http://wertel.eng.pro

 

[MadMango]

We have a company-wide intranet set-up. Engineering is the only department that updates information. I am the only one in Engineering to add data. HR still enjoys printing out flyers and fixing them to doors. Technology is great if used.

"Art without engineering is dreaming; Engineering without art is calculating."

Have you read faq731-376 to make the best use of Eng-Tips Forums?

 

[GregLocock]

KENAT - I think that's the way these systems always work. Those of us with elegantly naked heads are reluctant to commit our learning to an 'open' system since it gives the younguns an advantage that we didn't have.

Personally I think any normal company that doesn't encourage that sort of thing, or at least simple access to distributed learning, is doomed to fail. Case in point...



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[memiles]

I think I know what you are talking about.
Question: What size company? (Important)
Access files with Query?
Gain Access to file after?

If yes to last 2 let me know. The first is if the person I was thinking about will talk to you. (not me)

 

[code1]

Thank you all so far for all the advice. Seems like it's not going to be as easy as cataloging widgets. But at least that'll make it more interesting as it evolves.

Dear Memiles- to answer your questions:
1) Size of people- approximately 350 persons, with offices and jobs in various parts of the world- Americas, Europe and Asia.

2) Preliminary thoughts- possible to do keyword queries (but then will need "Owners" of the files to populate the fields. May not be able to search file contents, as I believe a large number of content will be scanned PDF files- for e.g. drawings, test charts, etc

3) Yes- will need to gain access to files after. This is the purpose of whole exercise- archive to build knowledge store house. Who has access/ can actually read- probably a management decision.

I remember that knowledge management was a very big thing a few years back (not sure if it still is) but it has died down in my part of the world now.

Now I need a workable file system first of all to catalogue information. How to manage the database (descriptions, key word searches, etc)- Excel, Access, etc, I'll consider after I have a clearer picture.

 

[MadMango]

You could arrange simple file folders by Projects, everyone remembers those. inside each Project folder you could have subfolders for Mechanical Systems, Electrical Systems, and Documents. Those 3 main subfolders could then contain other related subfolders, like drawings, contracts, design notes, POs, etc.

If you kept the same structure for all projects, people would know were to go. Everyone could have Read access, and only a limited number could have Write.

"Art without engineering is dreaming; Engineering without art is calculating."

Have you read faq731-376 to make the best use of Eng-Tips Forums?

 

[IRstuff]

That's what we have at the moment, and it's horrible exercise to find stuff. We have upwards of TEN network drives, each with thousands of folders.

Additionally, we have duplicated files all over the place. Similar projects wind up using similar components, but each project maintains their own components folders.

TTFN

FAQ731-376

 

[epoisses]

Yup, that's why it's more efficient to use a relational database rather than a simple "tree" structure.

The thing with knowledge management is, you need a briliant IT tech to set up the system, you need a brilliant engineer (or a team of) to determine the scope of what is to be stored and what not and to categorize, you need a brilliant manager to make people feed and use the system effectively. Once that is set up, you can do with less than brilliant people to do the scanning and all.

It is NOT a project that a small team of people can carry out during a couple of Friday afternoon meetings as item 39 on their to do lists. It is a MAJOR project that needs a dedicated team, very thorough preparation and very good marketing toward the rest of the organisation. As a simple rule of thumb, any engineering manager who cannot even manage to keep his own desk clean, should NOT be in charge.

 

[memiles]

Yep, epoisses
Precisely why I was asking for input from a very brilliant IT tech with an Engineering background.

Mike

 

[SomptingGuy]

For historical reasons we have two systems. Our library archives technical papers and they have a pretty good search engine (aided of course by the skills of those archiving the docs). For our general day-to-day business we have an intranet. Finding documents in that was virtually impossible (even if you knew their exact title!). However, we decided to pay for the Google serach engine and now it's brilliant.

 

[GregLocock]

Does the google search engine only work on text, or on pdfs as well?

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[SomptingGuy]

It does both. As I said, brilliant now.

 

[memiles]

Depends on how it was prepared/scan'd.

Sometimes the just the title, somtimes all. V8 is great.

 

[memiles]

add post

Acrobat V8 is great, so far