“We are not a software company.”
That simple phrase stated by a leader at Payette embodies an uncertain relationship that has developed within the AEC industry since the dawn of the digital age. The customized nature of design means that, inevitably, offices will create their own tools to test new, creative solutions. These tools may be as simple as a spreadsheet or as complex as a software plugin but they all raise a question that universally affects our practice: if our goal in the AEC industry is not to sell software, what is the fate of the software tools that we develop? Are they destined to remain on the company hard drive, collecting dust as artifacts of the designs that they informed? Are they meant to be reused and improved upon, in the same way that we iterate and improve upon our designs? Should such improvement and iteration be done only with in-house knowledge? Or should tools be shared with the industry, open to feedback, validated by expertise beyond that of the firm, and given a fair chance to change industry practices?
At Payette, we’ve discovered the value of sharing our tools and gathering feedback to improve them, which in turn, helps advance the profession.
For example, we originally devised the Payette Glazing and Winter Comfort Tool to inform a situation that impacted over half the projects in our portfolio: we began many of our designs with aspirations of using triple pane windows but, as project budgets were evaluated, strategies with long paybacks like triple pane were often the first to go. Soon after such decisions about the envelope were made, our engineers would inform us that we would need perimeter heating elements next to the facade in order to maintain occupant thermal comfort in winter. Intuitively, we knew that if our envelopes were well-insulated, our interior surfaces would be warm and we would have no need for this extra heating system. Yet there was little agreement among our engineers about what would qualify as a sufficiently insulated envelope or whether triple pane would be enough to meet thermal comfort standard without the heating elements. After a lengthy research effort to understand this situation, we realized that the ability to meet the comfort standard with triple pane depended upon the geometry of the windows. With punched or ribbon windows, triple pane was enough but, with an all-glass façade, it was questionable, and with multi-story glass, it’s out of the question. And all of these rules changed when one had a project in a warmer climate or if an occupant put on a sweater. To empower our practice to handle any situation that arose, we began developing a tool that could accept many inputs and run thermal comfort calculations to provide recommendations. We knew that we wanted to share it publicly on the web, but in the process of sharing, we discovered three unexpected benefits of doing this.
- TOOLS BECOME BETTER WHEN SHARED
“When an expert network is functioning at its best, the smartest person in the room is the room itself.”
David Weinberger, Too Big to Know: Rethinking Knowledge Now That the Facts Aren’t the Facts, Experts Are Everywhere, and the Smartest Person in the Room is the Room
It is rare for any one individual to have all of the knowledge needed to give the best answer to a question. Taking this idea to heart when we began assembling our tool, we realized that a lot of the code for calculating thermal comfort had already been written by the authors of the Berkeley Center for the Built Environment (CBE) Thermal Comfort Tool. The development of their open source code was done by leading experts in comfort science and was regularly maintained, so instead of translating the thermal comfort standards ourselves from the code books, we borrowed their code as a basis for our tool. This was the first time where we realized the benefits of building off shared work since it’s unlikely that our translation would have been nearly as up-to-date or comprehensive as with the CBE tool.
Yet this was just the beginning. Within a few months of completing the first version of the tool, we learned that the same team who made the CBE Tool had developed a new comfort model that was particularly relevant to the condition we were studying. Going through the rigor of surveying occupants under the controlled conditions of a climate chamber, the CBE scientists assembled a model to forecast discomfort from drafts at ankle-level, which was a lynchpin for understanding discomfort near cold facades. We replaced our previous model with their new state-of-the-art one and instantly realized that the results were more aligned with what we had seen in other studies. We finally had something that we were confident in and, while successful in its own right, this was just the beginning of many improvements that we made after receiving feedback from users outside our firm. Bug reports, feature requests, and questions on methodology all helped us realize a much better tool than we would have ever developed using only in-house feedback.
- PROJECTS BECOME BETTER WHEN INFORMED BY SHARED TOOLS
Having a good tool does not necessarily mean you will have good buildings. For our Glazing and Winter Comfort Tool, successful application required a consensus not just from Payette’s in-house building scientists but also the engineers with whom we work. Because the tool we built is open for anyone to experiment with, our engineers could validate it against their own methods, match their assumptions with ours, and ultimately sign off on our buildings with insulated envelopes and an absence of perimeter heat. It is largely for this reason that the fraction of projects using triple pane in our office skyrocketed over this past year. Had we not shared our tool, our consultants would understandably have been reluctant to sign off on our designs, effectively having to “trust us” that occupants were comfortable while they absorbed the responsibility. Getting past this critical barrier meant that we could finally ask an important question: which costs more – a triple pane façade without perimeter heat or a double pane one with it? For the two to three years since we started this process, the results have come out in favor of triple pane every time and, as a result, all of our designs are realizing a transition to more elegant, insulated (and cost-effective) facades. It’s unlikely that we would have ever achieved such a dramatic change in our building practices without the consensus built by sharing our tool.
III. INDUSTRY-WIDE PRACTICES IMPROVE WHEN INFORMED BY BROAD CONSENSUS
While some companies may view open source sharing as a threat to one’s competitive edge, in the case of the Glazing and Winter Comfort Tool, we haven’t found that to be true. The widespread use of the Tool validates the ideas and methodology behind it, allowing our firm to lead change in the way we design elegantly-detailed, insulated facades as a profession.
If we think back to our original quote, that we are not a software company, it’s true that our clients don’t hire us because of our software tools. Rather, we are hired because of the people in our office and our expertise. Ultimately, people design good buildings and, like any tool, digital or analog, the greatest software tool in the world is effectively worthless without someone to wield it. When you share a tool with the industry, you lose the right to sell that tool but you gain an enormous amount of experience and expertise through feedback.
Whether it’s through thoughtful reasoning or by accident, the vast majority of tools and scripts within the AEC profession remain closed and only accessible in-house. However, we have so much to gain by sharing them more broadly.
Contribution from Santiago Garay. The photos provided in this blog post are copyright property of © Payette,© Warren Jagger and © Keitaro Yoshioka