Drafting through the decades series
And we are back! In the last part of our series, we want to pay our respects to the CAD software that we use and reflect on what the future holds and what we envision as a sustainable solution for smart drafting software.
Part 5: In the aftermath of CADs
Computer-aided design improved continuously. In a few years it moved from a hard-to-use tool to being the official way of working. It became well established in every engineering office that wanted to work on projects with up-to-date quality and values. Especially AutoCAD, first released in 1983 for the IBM PC, marked up a milestone in 2D computer drafting.
Our first…
…was AutoCAD 12, released on 1992. We used it on DOS on a 386DX clocked at 33Mhz. If you installed everything it would take 10 of the 60 megabytes of the hard disk, and that was a lot of space at the time. Despite the complexities involved in fitting it in our workflow it was a great change of tremendous help, especially as demands increased.
Back in the day, software released every few years. Each new version was an actual experience and you had the time to incorporate any changes.
Five years but only two releases later, the improvement of AutoCAD was staggering. R14 was a mainstay of computer aided design. It even had real-time pan and zoom. Before that, the pan tool used to re-draw the design to the new coordinates you specified instead of moving it around smoothly like nowadays. R2000 had layouts. From then and during the last 18 years little has changed.
The attack of the clones
AutoCAD was not the only one though. The stagnation of CAD features among the versions and the exorbitant prices Autodesk was charging, led to the appearance of a multitude of clones. All this started when Softdesk created a secret AutoCAD clone in order to defend against potential competition by Autodesk. Autodesk acquired Softdesk but was not allowed to acquire IntelliCAD (the clone) due to fair competition rules. The source was sold to Visio, which was then acquired by Microsoft. Microsoft however did not pursue the AutoCAD clone business. In the meantime, another entity was formed that continued to license IntelliCAD code to multiple sublicensees who each develop their own branch.
However, all this tumult is nothing more than what CADs are from they very beginning; lines on a virtual paper. This needs to change.
Smart software
Software needs to be smart. Even though the word smart gets attached to almost anything nowadays, especially technology, it doesn’t always hit the spot. Building software for example is not there yet. Hopefully we don’t need to build it from scratch, just add sparks for innovation.
BIM (Building Information Modeling)
Communicating, cooperating and reviewing in an interactive visual environment goes a step further with Building information modeling. BIM is a 3D model-based process of working. Files are networked and shared to support decision-making in real time regarding almost any building asset. An engineering team is enabled to co-author the project at hand and coordinate everything during its life cycle until completion.
Building models should move on from plain objects to entities with attributes. Right now, we draw virtual lines of all the construction elements. But the software knows nothing about that. It cannot tell the difference between a slab and a column. As far as the software is concerned these are mere digital depictions. As a result, any modification must be done manually. For example, a bigger beam section means that you need to scale up all the beams manually. If we’re talking about evolution, the software should be able to understand and perform a beam scaling with a simple command such as scale up the beams by 0.2 and adjust everything, like the reinforcing bars, the column section, the concrete mixture ratio etc.
Objects need to move on to entities as well. If we draw a desk the software needs to know the concept of the desk. Hence scaling down the work surface shouldn’t result in thin desk legs unable to support themselves. When calculating the structural loads, the software would include the desk in the imposed loads.
Another task that should be done automatically due to entities is the tendering, both the itemization of the construction supplies and labor and their cost.
All these would improve the cooperation among the engineers as well. Let’s say that the architect wants to move a column but the new point is where electrical installations are. The software warns him, optimizes and proposes a new point. Consequently the electrical engineer avoids this problem. Thus the difficulty of collaborating with multiple partners is reduced.
The nature of an entity-smart software would give a real edge over traditional CAD design and its limited solutions.
Libre software and open standards
Up until now most engineering software has been proprietary, with minor exceptions. The industry-standard .dwg is not only proprietary but a closed standard that was never intended by its creator, Autodesk, to be used for sharing drawings (that is what .dxf files are for). IntelliCAD creators reverse-engineered it and created a rival proprietary piece of code that reads and writes .dwg files.
Thus it is almost impossible to create Libre CAD software to rival AutoCAD and it’s clones today. (The GNU project has been trying with LibreDWG which actually had another release recently, but it still ways behind.)
IFC is fortunately an open standard and thus support can be legally and easily added to Free Software projects. There are already open source libraries that read and write IFC files and work for supporting it in FreeCAD is underway.
Thus we hope that there will soon be a way to solve our engineering problems using exclusively next-generation libre software.
Version control
Extensively used in software engineering, version control systems records the life of a file. Additionally, by using project management tools revolving around versioning, you can have a complete documentation during every project, revert to a past state, set milestones, allocate responsibilities, track project activity. And that’s just the tip of the iceberg. For a better overview this article could be very helpful.
Our need for constant improvement, led us to integrate git to our working processes. It proved to be a major challenge, as any other digital transformation effort, and we are still struggling to find how to use the most out of it (local repositories, remote repositories and all that jazz). But we believe that is a very useful and powerful tool for us engineers.
It is the future, and we are not looking back!
OpenSCAD 3D modeling
Last but definitely not least is OpenSCAD. OpenSCAD is a free, open-source and script-only based modeler for creating solid 3D objects. It plays well with version control and allows the creation of fully parametric designs hence the model can be adjusted simply by configuring the parameters and modifying the variables. At the moment it is mostly used to design 3D printed parts and fixtures. However, we strongly believe that eventually it will influence the engineering world.
In the aftermath of CAD what else do you think comes next? What does the future hold? We do not know, but we surely want to keep exploring.
Thank you for reading!
