Kill Your Darlings Featured

My first touch with 3D/BIM was in 1997. After eight years of architectural design with pencil, Rotring pens, and some 2D CAD, 3D modeling changed everything. The architectural company I worked for at the time bought MicroStation Triforma — probably the most engineer-friendly software on the market: setting after setting, and a 3D window like a deep black hole where you modeled your objects. You learned that the depth lock was your best friend for controlling your moves, and that patience was a virtue while waiting overnight for 2D drawings to be generated from the model.

But somehow, I kind of liked it. It was logical and surprisingly creative to work with. The more you planned in advance how to build the model, the more you could benefit from the logic later in the process. After a while, I was able to plan, model, and manage huge projects very fast — from early sketches all the way to the site phase — using models, and this time not only the 3D geometry but also all the data in them.

In early 2006, I joined the BIM development team at Nordic Construction Company NCC — a step from design and architecture to “the dark side,” as my boss called it. The idea was to develop processes and tools to use BIM models effectively in all projects. It was an eye-opening experience. We had a front-row view of all units, operations, projects, tools, work methods, and models across all disciplines. Soon I realized the models were a mess. So the first idea was to create better guidelines for designers. That would solve it. Not. The second idea was to visit architectural offices and explain how and why. It helped, but it didn’t solve the core problem.

By 2010, BIM projects had become more common, but all of them still had major quality issues — especially data quality. It was an industry-level problem that people tried to tackle through national BIM guideline development projects. I was involved in RYM PRE/Model Nova, COBIM/YTV, buildingSMART/YTV, and the buildingSMART Talo. Cooperation with other BIM enthusiasts was great, but I had lost faith that new guidelines would solve this. There had to be another way.

Since 2006, we had tested, used, and implemented a huge number of BIM software tools to achieve the ultimate goal: how to take advantage of BIM models in a scalable way — in all units, in all operations, in all active projects. In 2010, we found Simplebim. We were thrilled. Finally, we were able to manage the models in a totally new way. The first use cases for Simplebim were trimming models for 3D site plans and for IDA ICE energy analyses. But it was just the start of something bigger.

Around 2011, I did my first model data enrichment tests with Simplebim templates. After years of trying to make modelers understand what different operations need from models, it felt liberating to realize we could enrich the models ourselves — in-house and by project teams themselves. That changed everything.

During the following years — first at NCC and later at SRV (2018→) — we tore down and rebuilt our whole BIM process multiple times. We minimized BIM guidelines and started developing IFC standardization and enrichment processes with Simplebim. Trying and failing, learning along the way — and disturbing Jiri and Sakari with lots of questions. We started with architectural models and scaled to all disciplines. First we handled models manually once a week, then automated the process with scripts. We tested the method with a few projects and ended up supporting around 100 projects per year with a fully automated standardizing and enriching process, handling more than 1,000 IFC models per month — with a BIM team of four and Simplebim.

Once we standardized IFC properties and enriched values, we were able to standardize user interfaces in downstream BIM tools for people in design management, site operations, quantity take-off, scheduling, etc. Standard user interfaces also made training more effective. The same core idea ran through the whole system: data should be easy to enrich and manage by the people who make decisions about it. And data should be easy to use and understand by the people who need it in projects — not only by someone who happens to know how to open an IFC file. Standard data helped a lot.

Another very important user group was management — people who have absolutely no time to open IFC files or learn how to search model data. But the data was highly valuable for decision-making; it was just in the wrong format. A standard data structure made it possible to export data from IFC and combine it in Power BI. And of course, we had the same data structure across all disciplines and modeling tools. That meant we could combine any model from any project and run any kind of analysis for management. Then we connected IFC data with other project data from other sources. That was cool.

Even though we used many different tools in our process, Simplebim was by far the most important one — the unicorn that couldn’t be replaced. But there was still plenty of room for improvement. If I really wanted to push this idea of “mass enrichment” as far as possible and give it a real chance, there was only one way to go.

In early 2022, I called Sakari and suggested we join forces and scale this mass-enrichment idea — to test how it would resonate globally. So I jumped again, this time from the building industry to the software industry — having no idea what I would face, but with strong faith this was the right move. The first mass-enrichment product was built for Simplebim Cloud as a full-service package. That didn’t work. The second product was also built for Simplebim Cloud, this time with Cloud Admin tools for managing the process. After several improvements, it started to gain interest. But the concept was still too complex.

Then came Dataflows. And that changed everything. Again. The mass-enrichment product was torn down and rebuilt as multiple smaller dataflow products, which you may now know as Installed Dataflows and Base Model. And they are just the beginning of something bigger.

I’ve killed my darlings throughout my career. Started over thousands of times. I don’t know a better way to develop — to keep improving, to learn from mistakes, and most of all, to learn from other people. And I know I’m a lucky girl — through all these years I’ve had the privilege of working with brilliant colleagues and customers. Still do.