CAD Manager-Document Archiving Strategies

Basic backups aren't enough—archive projects for future use.

The year was 1987 and I remember the sales pitch vividly: "Once your CAD archives are established, you'll be able to recycle entire projects in only minutes, and you'll never have to draw the same line twice again!" At the time it seemed like a realistic promise. In theory, it still does, even if the reality isn't nearly so rosy.

What about your company? Are you archiving your work in a manner that lets you recycle projects in minutes? This month I'll debunk some archiving myths and give you some ideas for tracking your CAD archives more efficiently.

Backups Are Not Archives

If you read nothing else in this column, please make sure you understand the difference between backups and archives. Backups are typically driven by IT (information technology), and the following characteristics usually apply:

• 1. Files are backed up only when they change.
• 2. Backups are usually captured each day, with a redundant weekly or monthly backup.
• 3. Backups capture files on network devices, but rarely do they capture work in progress on users' hard drives.
• 4. Backups are periodically stored at another site to prevent against loss due to fire, natural disaster or terrorism.

In most cases, backup data is stored on media such as digital tapes in a compressed file format. The media and compression factors conspire to make data restoration an IT department task because they have access to the backup tape drives and software applications. Furthermore, the backup process doesn't lend itself to finding files easily. For example, if you're trying to find the last version of a specific drawing, which backup tape do you think the file is on? In general, IT-driven backups should be considered disaster insurance that you hope you never have to use.

Conversely, archives include a total view of a project and all its data. In addition to archiving all data at the end of a project, a more detailed archiving schema could capture a project archive at various stages of the project's life to track its chronology. Expectations for an archive system are much higher than for a backup system.

Archive It All

When you archive a project, you may be tempted to think only of CAD files. With huge file sizes, external references, linked graphics and complex file-to-file relationships, CAD presents a formidable archiving challenge, to be sure. But concentrating only on CAD files for project archives neglects a large segment of total project data. In fact, non-CAD files tend to outnumber CAD files by a factor of four to one. Files that make up the rest of the project archive include:

• 1. E-mails
• 2. Spreadsheets
• 3. Word-processing documents
• 4. Transmittals
• 5. Vendor data sheets

I know many of you are thinking, "Why should I, as CAD manager, worry about all these other types of files?" My answer is that somebody must archive project data. If you don't do it, you can bet your CAD files will get messed up. If you're going to archive CAD data, why not put the other information in the archive as well? By doing so, you can earn some extra brownie points from management for taking charge, and you know the job is getting done right.

Metrics

Now that I've talked you into taking control of your archiving destiny, let's discuss what to archive, how often to archive it and how to document what you did. To jump-start the discussion, here are the minimum requirements of a good archiving plan:

• 1. Thorough capture of all data files, both CAD and otherwise, required to recreate the project at the exact time of archive.
• 2. The ability to document the files archived, their purposes and any special notes that a future project manager would need to understand the status of the project when it was archived.
• 3. Long-term storage for archive data at your business location and at a secure secondary location in case of total building loss at your main location.
• 4. A policy statement, or standard, for all project personnel that documents how active files should be stored so they can be archived properly.
• 5. An executive overview for senior management that explains how important accurate archiving is and why all personnel must follow archiving procedures.

As you can see, creating a great archive isn't just a matter of copying files onto a disk and throwing it in your desk drawer. Archiving is a comprehensive methodology that captures project data, knowledge and context within a standard framework that all employees must embrace. As with all things related to CAD management, you must keep management informed, so do a good job on the executive overview.

Archiving Formats

In years gone by, the archiving format of choice was the hard-copy print or original drawing vellum. As drafting boards have disappeared, we've been forced to decide what sort of digital format to use to archive our information. A few questions illustrate the archival format conundrum very well:

• 1. Should you archive CAD data in native format and replot later if needed?
• 2. Should you archive CAD data along with a plot file so you'll know what the printed documents looked like?
• 3. Should you scan in as-built or markup prints or retain them as hard copies?
• 4. If you do scan, what format should you store them in?
• 5. Should text-based documents be stored in their native applications or Acrobat PDF, or both?

These concerns should be considered very carefully with an eye toward the ever-present chance that file formats will change over time. In fact, software applications can be bought, sold or made obsolete well before you ever need to access your archives.

To mitigate the risk of data obsolescence, I suggest the following archiving practices:

• 1. Archive your native CAD files and the plot files. I realize that tracking plot files is another burden, but years from now you may not be able to reproduce the exact plotting environment you're using today, so repeating the last known plot may become impossible over time. Safe formats for storage are an HPGL-based plot file, Autodesk's DWF or Adobe's PDF.
• 2. If you can't manage plot files, keep a set of record prints for scanning. Scanning files is much more labor intensive than creating a plot file, so this option is a fall-back position.
• 3. Archive the entire directory structure so as not to break paths to related files such as xrefs, fonts and images. In fact, you may want to note which drive letters were in use during the time of archive should you need to exactly replicate the drive and path structures later. I've seen too many cases where xrefs had hard-coded paths that wouldn't resolve after restoring from archives.
• 4. Store scanned documents in an industry-standard TIFF file format. Make sure it can be opened in an inexpensive utility such as Windows Imaging or Paint. Beware hybrid or custom TIFF file formats that can't be opened without special software.
• 5. Save word-processing documents in the native format and in Adobe PDF. My logic here is the same as storing a plot file to preserve the visual formatting of the document without worrying about errant pagination or margins.
• 6. Store scanned paper documents, such as contracts, in PDF format.

You may take issue with my insistence on archiving information in native and visual formats such as PDF, but my reasoning is that when archives are searched years from now, a visual search methodology can use the PDF files easily, without having to install old software applications to open the native data. And while I can't promise the PDF format will always be valid, I think it's a safe bet.

Media

When creating your archives, the media you use is of real concern. I'm sold on writeable DVD drives because they are as stable as CD-ROMs, have much higher data densities and enjoy such widespread use that continued support of the media is assured. DVDs are sturdier than tapes and don't require any special software to navigate their contents. As with any media, it's doubtful that DVDs would survive a total fire loss even when protected in a well-insulated media storage safe. Therefore, an extra set of archive DVDs should always be kept at another location as additional insurance.

Revisit and Refresh

Just because you've completed a project archive doesn't mean that you're done forever. Periodically, you'll need to re-examine your archives to make sure that the file formats you stored can still be used. Let's envision the nightmare scenario of your CAD vendor going out of business and sticking you with thousands of files that will become obsolete. In this case, you'll need to investigate what will be required to translate all your old project archives to a new CAD platform immediately while the technical resources to deal with the translations are available. Don't believe me? Just try getting a few thousand CADvance or I-DEAS files converted to AutoCAD right now!

This nightmare scenario certainly isn't typical, but it's real. I've seen it happen to several of my customers over the years. There's no formula I can give you for this process other than to stay on top of the technical factors that governed your CAD installation and react accordingly.

Archives Live Forever

Very few things you do as CAD manager live on as long as your archives. Because CAD data is so rich and has such complex file-to-file relationships, it behooves you to make your archives as bulletproof as you can. Use this article as your checklist to start an archiving program or revamp your existing one. Either way, your company will be better served by the creation of high-quality archives.

Robert Green performs CAD programming and consulting throughout the United States and Canada. Reach him at rgreen@greenconsulting.com.