Robot beetles build the future?

The Smart Geometry 2016 conference was held at Chalmers in April. It was the 13th time that a diverse group of architects, researchers, engineers, hackers and like-minded people gathered around the question of how digital technology is transforming architecture. Petra Jenning and Edvin Bylander from FOJAB arkitekter participated.

The theme of this year's conference was Hybrid Domains, and the focus was on the interdisciplinary and transdisciplinary, with many examples of how fields adjacent to architecture work with computational design and digital manufacturing.

One of 10 workshops was held by Petra Jenning, Head of Computational Design at FOJAB arkitekter, together with Kirstin Peterson from the Max Planck Institute, Stuttgart and Nils Napp from the University at Buffalo, New York.

- In our workshop we wanted to explore how we can make very simple robots, agents, interact by just changing their physical body and the environment they move in. The robots only have the ability to move in a straight line, but by giving them different shapes, they interact in different ways with each other and things around them. They build together structures with controllable properties, even though they have no intelligence or understanding of what they are doing," says Petra Jenning.

The very concept of getting several simple robots to build complex structures and adapt to reality was something that several speakers highlighted as the most likely way to bring robots into the construction process. Foster + Partners' proposal for NASA on how to build on Mars was one example.

- The aim is to better understand how simple rules and relationships can create complex systems with predictable properties without describing the final form. "This is how physical structures come about in nature, whether it is a skeleton in our body or a termite stack," says Petra Jenning. "This is exactly the opposite of how we have conventionally constructed things from the Renaissance onwards.

Petra explains that there are two basic things in this: One is performance. You are interested in the characteristics of the built object, not the form for the sake of the form - here we use the concept of performativity. The conventional drawing describes how something looks, not what effect it has. Now we are designing a process that leads to a finished object.

The second is deployment. We strive for technical solutions that are robust and agile enough to work in real life. There is a problem about how advanced robots can work in practice in an unpredictable and messy environment such as a construction site. Problems related to robot-human safety, range, cost, lack of robustness. Autonomous distributed systems, consisting of several simpler smaller robots working together, can be robust, agile and better suited to real-life situations.

Robot researcher Kasper Stoy from the IT University of Copenhagen called for this as a next step in the development of computational design. "We have begun to link digital design with digital manufacturing of building components, but what is missing is the most critical part, the actual construction. Not much has happened here since the invention of the digging bucket and the crane. To find new solutions, architects need to be visionary and communicate their ideas to robot builders.

- There is much to learn from how related industries work with innovation in this field. Not only in what they do, but also how they organize themselves and what their processes look like. For us, Smart Geometry 2016 is a confirmation that we are on the right track and have a relevant approach, says Edvin Bylander lab director at FOJAB arkitekter.