I had a chance to read Matt Lombard’s – What if you CAD data were in a database? The discussion around this post made me think more about CAD data organizations, CAD-PDM development, new technologies and some trends such as cloud and noSQL. The initial triggers for the conversations came from some new products, releases and announcements by CAD/PLM vendors over the past years. Autodesk, Dassault and SolidWorks were first to push “cloud race” start button. So, the question about what will happen with one of the biggest assets in manufacturing – CAD data, getting more and more priorities in discussions. The introduction of Dassault V6 put an additional priority. DS made their decision to bundle CATIA and ENOVIA V6 database. By doing so, Dassault opens a new round in the development of CAD-PDM dependencies.
Files, Desktop, Cloud
One of the most fundamental abstraction computer software holds for almost 4 decades is the abstraction of a computer file. Combined with desktop metaphor, developed by Windows and Mac software, it used by the majority of users today. During the last ten years, the internet development created some ideas for how post-desktop world may look like. It is still a big question if all “beyond desktop” stories can make a reality. There is a great read about a future of post-desktop world, I can recommend you – Design Integrated Digital Work Environment. At the same time, the simplicity of “file-based” metaphor and large amount of existing application and CAD data can make vendors to duplicate this abstraction level on the cloud. Remember – simplicity always wins!
CAD was a first tool that went to mainstream in manufacturing organizations. Obviously, CAD application, as a desktop tool started to work with file. Later, the issue of controlling files and managing of CAD data came into the game. PDM served needs of the customers to organize CAD data storage. PDM can remove the need of file management. I just discussed this topic on my blog few days ago – Why do we struggle with file names in PDM/PLM? However, at the same time, the complexity of CAD/PDM interfaces and implementation cost slowed down distribution of PDM tools. The latest trend I discovered is that PDM interfaces have a tendency to be standardized by CAD manufacturers. The next expected step in this direction is to make PDM a standard functionality of CAD system. The real showstopper in this direction is a complexity. Many of the users today are continuing using CAD tools and see PDM as something that only makes their work more complicated. The future step in achievement of simplicity is to bind PDM into CAD and make them inseparable. There are advantages and disadvantages. It can simplify the tool and user interface. At the same time, usage of multiple CAD systems and other customer’s reality are stopping customers from implementing these strategies.
There are few interesting trends I want to mention. Last ten years of the internet introduced additional technologies in the data management space. Relational databases are still the data management mainstream. However, thinking about noSQL solutions, graph databases, proprietary data stores like Google Big Table, Open Source solutions and others, created some opportunity to change “SQL-only” paradigm used by PDM/PLM systems. I can see two most important parameters that will drive cloud-based-data technologies. Scale and Cost of Change. “File” is a well-understood user paradigm and companies can continue to use it even when the cloud will use different storage technologies.
What happens on the side of vendors? I think, CAD/PLM vendors started to watch space more closely. Autodesk experimented with cloud technologies and introduced few products (including acquisition of Israel-based Visual Tao). Dassault announced about their plans to bring V6 platform to the cloud.
Even if the cloud-y future looks very promising, the question of data lock-in is an important and requires special attention. There are few issues related to the lock-in. As soon as you have your CAD data in files, you physically own it. So, even if you migrate to another CAD system, it is up to you (and translation tools, of course) to move it to another system. However, when you need to pull your data from cloud it can be different. For example, try to pull your emails from Gmail. Google watches very carefully large data volume migrations. We don’t know yet what path CAD companies will choose in their decision to move your CAD data on the cloud. What if you will be licensing your traffic volume? C’mon… It can be just crazy.
What is my conclusion? File-paradigm was born as a technology. However, for the moment, it is just a matter of usability and portability. The internet creates new paradigms that will proliferate into engineering system space. The dilemma to store CAD data in a file vs. database will become an issue of technology. Just my thoughts…