Is Your Data Management Strategy in Sync with CAD Needs?

CAD Manager Column: Do your users rely on free file-hosting services like OneDrive or Dropbox to sync their files? It’s time to get educated about the requirements — and real risks — of this approach.

I recently received a question from a CAD manager asking for my opinion on how to deal with CAD/BIM users saving their work in file-hosting services such as OneDrive or Dropbox. My immediate reply to her was to run away from these options as fast as she could, and embrace true data management control instead. But upon further discussion, I realized I needed to address multiple data management concepts and justify my fear of using this type of tool as a poor man’s data-control scheme.

In hopes of helping the many CAD managers who are struggling with data management issues, I’ll share with you a distilled version of our conversation on the topic. Here goes.

The Elements of Data Control

Whether you’re dealing with CAD files, building information modeling (BIM) models, spreadsheets, or PDF files, the basic mechanics of managing files are the same. They include the following:

Data security. The surety that files won’t disappear, be accidentally overwritten, or vanish into a cloud-based black hole.

Simple, dependable revision control. The ability to trace the development of a file over its lifetime. What did a drawing look like at Rev 1, Rev 2, etc.? Can you reliably get back to an old revision if you need to?

Mechanical/BIM/Civil revision control. Similar to the revision control issue above, but more complicated. These software tools have complex interfile relationships such as parts in assemblies, coordinated systems inside buildings, point clouds and surfaces along with in-ground infrastructure, etc. Change a part, and the assembly and drawings change; change a building size, and the HVAC ducting has to change. Can you keep track of all these relationships and revisions?

Parallel revisioning. This refers to the possibility that two people — usually in different branch offices — may be editing the same file at the same time. This “whoever saves last wins” work management methodology creates uncertainty at best, and rework in most cases.

WAN topologies. Wide-area networks (WANs) allow work teams all over the country (or world) to share files and work responsibilities, which only amplifies the types of problems already outlined. To make matters worse, WANs can be very slow, making an already difficult coordination project unbearable when large files take many minutes to load/update over slow WAN connections.

Ever wonder why it is so hard to control project files? It’s because there are many things to keep track of — and many ways to lose track of key data. In short, data management is difficult because so many things can happen that allow errors to creep into the process.

Sync Architecture

Before we dive into what syncing files really means, let’s take a few moments to define the basics. Imagine that you have a desktop machine and a laptop that you’d like to keep perfectly in sync, with all your files exactly the same on both machines. This is the architecture that programs such as OneDrive and Dropbox create so that you can edit a file on either of your machines and those changes will be replicated — or synced — to the other machine. The key things to recognize about this data synchronization architecture are:

A solid Internet connection is required for both machines. In addition, the speed of that connection is important because it controls how quickly changes can be synced between the machines.

Offline machines put syncing on hold. If you take your laptop on a trip to a remote job site with no Internet coverage, then the changes you make can’t be synced to your desktop machine until your laptop hooks up to the Internet.
 
Data loss can happen very easily. What if your laptop gets stolen while you’re visiting that job site?

Synchronization can be lost very easily. What if your desktop machine locked up before it could sync everything? Now the most recent versions of the file will not be available to your laptop, whether you have Internet connectivity or not.

These services depend on a vendor’s cloud architecture, security, and app reliability. Your files will only be as secure as the vendor’s cloud infrastructure is. While OneDrive and Dropbox are certainly great programs, they’ve experienced glitches in the past, and likely will again.

In summary, while these applications have a solid theoretical basis of operation, they only work within a certain set of conditions — and they present risks that could affect data security.

Single User vs. Corporate

Now that we’ve covered the basics of a sync architecture, it is time to think about the scale of the architecture. It is one thing to utilize services such as OneDrive or Dropbox for your own personal use, but quite another to put an entire group of company users on them. Why? Here are my core reasons:

Scale of possible loss. If I lose some data, then only I suffer, and rework is limited only to my own work product. But if my entire group loses data — or works with bad data — the magnitude of the rework goes up astronomically. Go back through the data loss and sync loss scenarios above, and imagine them on the scale of an entire design team.

Lack of revision control. Saving in a OneDrive or Dropbox architecture gives me no revision control and no real capability to roll back file versions. While the latest saved file will be synced (unless problems occur), there is no way to go back to a given version of a file a week ago, month ago, or at a major project submittal last December, as an example. The simple fact is that revision control is a key aspect of file storage, and simply syncing files doesn’t mean you have control of revisions.

File collision handling. What happens when a file can’t sync right away (when someone makes edits during a flight or on a train, for example), and those changes are not interpreted by the system until hours later — or even until the next day, when the user finally syncs up via the Internet? If you've ever had a scenario like this in your own day-to-day work, you know it is extremely easy to end up with multiple named versions of a file — and entirely possible to lose work if all the syncing doesn't go exactly right.

Are Business-Focused Apps the Answer?

To address the types of concerns stated above, vendors including Microsoft and Dropbox have beefed up their group file-handling services via applications such as SharePoint and Dropbox Business, respectively. The logic goes that if your CAD application integrates with one of these options, which have more robust file-sharing features, corruptions won't happen. And while this might be true, nasty little issues such as bandwidth, latency, and high network traffic can still conspire to create slow, clunky, and error-prone experiences.

In addition, each CAD vendor has to provide its own integration to these types of tools to ensure that file locking and saving are properly controlled. Think about submodels in Revit, assemblies in mechanical modelers, or xrefs in CAD applications, and you’ll realize that the degree of difficulty is now approaching critical mass. Not only can a single file become a problem, but the relationship between files as they are edited and synced must be managed as well. Now throw in some proprietary file types that have a wide variety of locking and open/close architectures from the main CAD/BIM program, and you've got the recipe for a real mess.

The Right Tool for an Essential Job

So why the dissertation on how all this could go wrong? Because in my experience, sooner or later, it will.

Simply put, don’t be fooled into thinking that free file-syncing tools are a viable solution for managing your CAD data in a company environment. To do data management correctly, you’re going to need tools that are designed, scaled, and tested to operate in large CAD environments — and we’ll be looking at those tools in an upcoming issue of the CAD Manager’s Newsletter.

Summing Up

I hope that this article has provided a wake-up call on the topic of data management, and file-sharing utilities in particular. Remember that your first priorities as CAD manager are to facilitate the production of work and protect your data, so taking data management lightly simply isn’t an option. Until next time.