Friday, September 5, 2008

BIM From Autodesk - The Promise Past

This post is transferred from my old blog, the CADRE BSD Blogspot.

This year we celebrate the 25th year of Autodesk - the biggest and best CAD programmers on the market. But how did it get here? What is it that made Autodesk into the powerhouse it is today? The AutoCAD Way You've probably heard a million reasons why, but to me it boils down to a simple thing. The AutoCAD product line had the easiest and most open interface available in the market for years - in other words, whatever you wanted to make of it. The user could (and did) adjust how the program worked to suit a specific need.

When taken into the context of changing the design industry from working on a drafting table to producing electronic documentation, there wasn't anything better on the market. So, the massive learning curve really occurred in the late 1980's into the 90's, as users were not only learning CAD but learning to use Windows and computers as well. For many years, the more ambitious users began to write their own programs to run on top of AutoCAD. One of the early programmers, Dave Arnold, was writing applications for civil engineering. He saw the potential in another program, ASG, which was designed to work with architects. He purchased this program and created Softdesk, the first AutoCAD add-on package that encompassed the entire AEC design line - and did it successfully.

Enter our first promise - a CD called the AEC-X solution was produced, demonstrating how a door could be placed in a wall in a CAD file. The door automatically cut an opening in the wall, and updated a nearby schedule. From the schedule, a user adjusted the size of the door and the door resized itself automatically. This first example of object based CAD got my motor running - in a meeting with a few of my friends that were working at other firms at the time, we all agreed that if this technology really worked, the many late nights we spent writing lisp, menus, etc. would be a thing of the past - we simply wouldn't be able to write code to reproduce what Softdesk was proposing for what the price was going to be. The problem was the technology wasn't developed - but the idea was enough that Autodesk decided to purchase Softdesk - and begin to develop this object-based technology.

Enter Architectural Desktop (now finally called AutoCAD Architecture - the right name that should have been used in the first place) - one of the first object based CAD packages that included physical representations of building components that could be used in either a 2D or 3D form. This is where my first red flag popped up - it was great to have an architectural tool, but the engineering tools would also have to be developed - a process that started 10 years ago and continues today with the AutoCAD MEP product line.

This all occurred right as I was coming into the Autodesk reseller channel - a switch I made based on a conversation that I had with my employers at the time. I had written (and helped edit and write) a series of lisp routines, menus and scripts that helped to create telecommunication towers for the next generation digital phone systems we all use today. After winning an award for improving the work process, I asked for a position that would allow me to do this for the entire company - to which the response was, we won't have a need for that kind of position (???) - a job that allowed me to help us improve our whole work process.

So, I left - for a job with CADRE, which would allow me to take what I'd learned over the years and use it to try to help others improve the quality of other designer's lives. I wasn't always successful, but the ability to fulfill a promise to myself was there - and is still motivating me today. But ultimately it was the promise of the technology that was really motivating me - the potential of object-based design offered a new way to do what I had spent many long nights at work doing, away from family and life - this was an opportunity to improve that quality of life.

Getting back to the programs - now that we had Architectural Desktop, we set out to prove to the architects that there was a better way to do things - and failed miserably. The problem we encountered had to do with the fact that we were positioning ADT as a 3D program from start to finish - something the architects didn't immediately embrace. We started off telling people to use the massing tools or space tools to create the schematics,and then convert these to walls, doors, windows, etc. Most architects were still doing this on trash paper - and they weren't the ones using CAD. We put them into a dramatically uncomfortable area - the older architects were not necessarily the ones actually using the computer. The industry, even though it had adopted 2D CAD readily, had not really changed its workflow - an architect still handed off most of the drafting work to drafters, interns and non-degreed designers. 3D meant they had to start using the tools - and many felt that was beneath them.

So we had to change gears - we started approaching ADT as a tool to improve drafting, and for a while walked away from one of the basic premises of the program - that by using 3D we could visualize and improve how the building was designed. The tool had (and still has) tools to make this easy, but we simply didn't present it correctly. Over the next few years, we made little headway, until the AutoCAD 2002-based Architectural Desktop 3.3 was released, finally containing all of the basic components of a building that could be made as a 3D component. While users still weren't using all of the tools available, we began to see a shift. The promise was beginning...

Enter the Revit...

Around this time several programmers from a competitor to Autodesk left to form their own CAD package - but from a manufacturing perspective. The idea was that a building was nothing more than the same thing manufacturers were producing - except on a much larger assembly. They created Revit - a parametric approach to building design. It offered the same tools as ADT - walls, doors and windows - but instead of producing multiple drawings, a building was treated as a single file and project. You would look at different views of the model, and use those to produce the details, sections and elevations. The initial problem was that it didn't follow the traditional workflow. Originally only one person could work on the model at a time, and until worksets were added to the program, the workflow shift was too great for most firms to swallow (remember - most of the designers didn't do their own CAD at this point).

But the marketing staff at Revit targeted the principals anyway, emphasizing the fact that the interface was simple - and couldn't be edited the same way AutoCAD could, so it would be more consistent. One of my coworkers and I went to Waltham, Massachusetts to get trained on this product - and drank the koolaid. The promise of parametric-based object technology offered many improvements to the AutoCAD process, but there was also many red flags for me - right back to the same old one - no MEP or structural engineering tools. Initially it would be a hard sell, but Autodesk came knocking and purchased the program...and the problems began.

Mom always likes you best!

Autodesk now had a dilemma - they had two building design packages - with the new kid running around telling everyone that Autodesk was going to kick the other child out of the house(anyone who has kids knows exactly what I'm talking about).

Herein lies the problem - what in the world do you do with two powerful design programs? They couldn't walk away from Architectural Desktop and Building Systems - the amount of money spent developing and improving these programs was immense, and the user base had grown dramatically. While the AutoCAD based products weren't perfect, the feature set was more mature - and Autodesk could not financially wait until Revit MEP and Structure products were on the market and able to do what needed to be done. So they did what they should have done - continue to develop both product lines.

But we began to see something new - there were plenty of architects that had not bought into the Architectural Desktop way of doing things. Even though it was better than the line based CAD they were producing in other products (including plain AutoCAD and AutoCAD LT), there was a resentment against Autodesk, since it was the largest package. Many of the non-users and principals were only aware of how AutoCAD worked - which didn't include purpose built tools for the building design industry, so they weren't necessarily comparing apples to apples.

Enter the Revit - it was easy to use, had the tools, but most importantly, was NOT AutoCAD - even though it was owned by the same company! So the Revit rush began... and now we're in the present...but we'll save this for the next segment...

thanks - David B.

No comments: