From interviews with experienced breakwater designers it appeared that in the current breakwater design process not all feasible concepts can be explored due to time constraints. Furthermore, also conservative assumptions have an influence on the design process. Both these factors make a suboptimal design more likely, which can result in losing a tender. To make an optimal design more likely more concepts have to be explored in the early design phases, as these phases have the most influence on the eventual concept (Rustell, 2016).
One way to explore more concepts within the same timeframe is to use a design automation tool. However, to use a design automation tool, a design automation tool must be available. This was the problem I encountered during my research, as several design automation tools exist, but unfortunately none of these tools are available as open-source tools. This is precisely the challenge of the construction industry, instead of using and further improving an existing tool, I needed to develop my own tool from scratch and thus preventing the continuous improvement of tools.
The main challenge: open-source the tools!
I understand that for companies it is difficult to share their design tools, especially when the tool helped them to win a tender. Sharing the tool is then perceived as giving your competitors the same advantage as you have. Therefore I will not focus on companies. No, instead I will focus on students and universities because, their main goal is not to win tenders. On the contrary, their main goal is to progress and develop knowledge.
Therefore, the best place to start open-sourcing tools are the universities and applied universities, as these institutions are based on democratising knowledge. When the students and researchers from these institutions start sharing their code and tools under an open-source license we can compound their impact and start a process of continuous improvement.
Existing tools will then become digital building blocks for the development of new tools. Universities should therefore encourage students to open-source their code and tools. By, for instance, sharing their tools on GitHub, or alternatively, on a dedicated GitHub for engineers. This will enable the whole engineering community to use their tools. When engineers are then using these tools in their design processes they can provide the students and universities with feedback, which can be implemented in the design tool, starting a process of continuous improvement.
The breakwater tool as an example
So my breakwater tool is not the first tool, but it is the first tool available under an open-source license (https://github.com/Sander-w/breakwater/). The main design philosophy behind the tool was based on Lawson and Dorst (2005) who explained the importance of experience in a design process, as experienced engineers intuitively and immediately performs the necessary design steps. This statement was also confirmed during the interviews I conducted, as the interviewees frequently explained their design process by using examples from other projects.
This is an important part of my tool, as the goal of the design automation tool is not to replace the engineer. On the contrary, the design automation tool must mitigate the weak point of the engineers, which is quickly and repeatedly performing the same computations without mistakes. On the other hand, the engineers have something the tool does not have; their experience and engineering judgement. By integrating a design automation tool in the design process we get an integrated design process where the main strengths of both the tool and engineer are used.
However, my tool is a purely Python based tool and this limits the user base, since not all engineers are skilled, or comfortable, enough to use a Python tool. This raises the threshold for using the tool, which is undesirable as we want to improve the breakwater design process by lowering the threshold as much as possible. Therefore, an easier to use interface is needed, for instance a graphical user interface. On step further is to create a web application out of the design tool. The tool can then be accessed through a web browser. Besides not requiring a software install, it will also make the tool available from all location, as long as you have an internet connection of course.
But, creating a web application is a tedious job and requires a lot of time, time you might not prepared to invest as the design tool was the end goal. To bridge this gap VIKTOR offers a Python SDK that can be used to create a web application. With this web application engineers who are not familiar, or comfortable, with Python can design breakwater concepts in a matter of seconds. This greatly expands the user base, as everybody can use a website to enter the required values, knowledge of Python is no longer a requirement!
The design automation tool together with the VIKTOR application can be used to quickly design breakwater concepts. This allows engineers to investigate the influence parameters on the design and cost, and thus explore more diverse concepts in the same amount of time. Because, more concepts are explored by engineers it is credible to assume that the chances of finding the optimal design increase. As a result, the chances of winning the tender also increase.
So back to the main challenge: to open-source the tools. My research showed that an open-source tool can help engineers in their design process, and thus increase the chances of winning a tender. Unfortunately, students forget or do not want to share their tool after they graduated. However, students are in possession of the missing link, the design automation tools! So remove all the dust from the tool you created during your graduation, or research, and upload it to GitHub under an open-source license. Democratise your knowledge so that the construction industry can improve!
And remember, don’t be afraid to share your tools or that someone will find a bug in your code. The biggest compliment you can get is someone else, or a whole company, using your tool!
Author: Sander Winkel
Bernal, M., Haymaker, J. R., and Eastman, C. (2015). On the role of computational support for designers in action. Design Studies, 41:163–182.
Cederfeldt, M. (2007). Planning Design Automation: A Structured Method and Supporting Tools. PhD thesis, Chalmers University of Technology, Göteborg, Sweden.
Lawson, B. and Dorst, K. (2005). Computational and Cognitive Models of Creative Design VI, chapter
Rustell, M. (2016). Knowledge extraction and the development of a decision support system for the conceptual design of liquefied natural gas terminals under risk and uncertainty. PhD thesis, University of Surrey.
Winkel, S. (2020). Improving the breakwater design process by using a design automation tool. MSc. Thesis, Delft University of Technology. http://resolver.tudelft.nl/uuid:494fa652-2125-4352-aeba-041eebcefde9