Architecture by Robots, For Humanity

Courtesy of – ROB/Arch Workshop, Rotterdam

Architecture is quickly adopting the popular technology of robots. Although it is slightly hard to define what “” really means, for architecture, it tends to refer to anything from arms to CNC mills to 3D printers. Basically, they are programmable, mechanical, and automated instruments that assist in processes of digital fabrication.

So, what might robots mean for architecture? A more precise architecture which could contribute to a more sustainable building life cycle? More innovative design derived from algorithmic processes? A more efficient prefabrication process that could reduce the time and cost of construction?

Probably a mix of all three. But more importantly, what might robots mean for humans? Robotic replacement for the construction worker? Loss of local craftsmanship and construction knowledge? Maybe. But I might reformulate the question. Asking what robots mean for humans implies passivity.

What I ask, then, is what can robots do for humans?

Courtesy of ETH Zurich – “The Programmed Wall”

Robotic fabrication methods are very efficient. They can do what most human labor can with equal to more precision, for the same amount of money, in much less time, and without concern for injury or labor laws (e.g. the work week). If need be, they can work for twenty-four hours a day, seven days a week. Quantitatively, that means a robot can work over four times longer per week than the average worker. And you also don’t have to pay a robot, they work for free (minus the energy bill).

If the advantage of robots is their efficiency, why aren’t we using them for things that require cheap and fast construction (and by require, I mean in the interest of livelihood, not budget constraint)? This includes anything from emergency structures for disaster relief efforts and refugee camps to temporary structures for the homeless. Surely it is the market capitalist economy that drives the development of most technologies, including robots. But doesn’t the market capital model of getting more bang for your buck fit perfectly into efforts that rely on making a lot (shelters, facilities, etc.) out of a little (time, money, and materials)? Imagined thus, robotic efficiency can do more than just save people money, it can make the act of aid more economically feasible.

So how can architects practically apply these technologies to structures of aid?

Courtesy of The Atlantic – Hurricane Sandy Aftermath

Prefabrication, in which buildings are either assembled onsite or prefabricated and brought to the site whole, dramatically reducing or even eliminating on-site construction time, could easily gain from constantly working robots. By enabling a more efficient work cycle, robots pose great potential for improving the speed of the prefabrication process.

Moreover, prefabrication is crucial in disaster stricken areas or refugee camps, where on-site construction is detrimental/impossible and importing whole structures, parts, or even materials, can be complicated and expensive. So, what if we put robots on the ground and fabricated structures at the disaster site? Or at least nearby? Using local materials? With very little human assistance? Not only could this reduce cost, it would involve locality and enable alternative energy uses. Emergency shelters could be mass-produced in less time and for less money, thereby increasing response time and reducing construction costs. More time and money, then, could be spent elsewhere, on things like food and water.

Courtesy of ETH Zurich – “R-O-B Mobile Fabrication Unit”

3D printers, too, make their case for onsite utility. Behrokh Khoshnevis, a USC engineering professor, recently developed a prototype 3D printer that could construct a building onsite in twenty hours. Known as “contour crafting”, the process essentially builds up the walls of the house in layers of a composite concrete capable of supporting itself during construction. According to his website, Khoshnevis imagines this technology benefiting victims of disaster and war:

Contour Crafting technology can deliver strong dignified houses to disaster victims very rapidly. Construction by Contour Crafting can build a 2,000 square foot house with all utilities for electrical and plumbing in less than 24 hours. Contour Crafting technology is adaptable and can use in situ construction material, thus eliminating the need to transport materials long distances, saving the time and costs associated with transportation. Since Contour Crafting is an automated process, labor needs are highly minimized allowing relief workers to allocate their time and effort to rebuilding local infrastructure such as water sanitation and distribution systems, roads, electrical and communication systems as well as irrigations systems. In this way Contour Crafting has the potential of providing disaster survivors not only with dignified shelter, but also with more resources to rebuild their lives and their communities.

YouTube Preview Image

Although this technology ultimately remains unproven, it marks a promising effort for implementing an architectural robotic process of aid. (See our previous coverage of this, and more possibilites here.)

Robots might also be able to participate in the initial response to disaster and destruction by contributing to debris cleanup. Although at the moment robot arms have limited abilities to work onsite, in a random and unpredictable environment, Testa & Weiser of Sci-Arc believe that robots will gain enough sophistication to work under complex site conditions. If they are right, and robot arms learn to adapt in unfamiliar conditions and function onsite, they could improve both the speed and safety of disaster response by automating the often slow and dangerous task of initial cleanup for rebuilding.

Courtesy of Sci/Arc – Sci/Arc Robot House

We, Americans, are notoriously bad at disaster relief. Hurricanes Katrina and Sandy have left many people homeless, and viable solutions to quickly shelter them have been slim. Why? Why aren’t we 3D printing shelters for the homeless? Why, instead of prototyping different algorithmic forms for the sake of experimentation, aren’t architecture schools prototyping emergency shelters for disaster victims? The paradigm of experimentation must shift from exploring a technology for its own advancement to exploring a technology for the advancement of humanity.

We should consider robots as instrumental to architectural and human aims, not generative of them. Architects should not design buildings for the sake of being technological, or even in the interest of technological formulation or production. They should design as they have always designed, and employ the technologies available to them to enhance their process of creation. That is, robots should not inspire architecture for the sake of robots, they should simply assist in architecture for the sake of people. Hastily building emergency shelters with robots achieves not just a technological goal of faster production, but a human end in the safety of disaster victims. After all, if robots do in fact pose a potential to increase our national and individual capacity to aid in the (currently deficient) process of relief, we owe it to our society – and to architecture – to pursue those potentials wholeheartedly.

Cite: C. Molloy, Jonathan. "Architecture by Robots, For Humanity" 29 Mar 2013. ArchDaily. Accessed 27 May 2015. <>
  • Pingback: Architecture by Robots, For Humanity | low

  • Jeph

    I think full automation also has the potential to bypass the need for architects in the future. And ultimately, architectural design will be democratized the same way people are talking about democratizing manufacturing.

  • Kieran Reichert

    Very interesting perspective, Mr. Molloy. I wonder what you make of the human costs of turning to robotic construction, in terms of loss of employment. has close to 5.5 million people in the U.S. working construction, and therefore implementing robotic construction could cost at least some of them their jobs.

    How can we balance our limitless drive for technology with the unchanging fragility of humanity’s survival?

  • Jonathan C. Molloy

    Kieran: your question is an extremely important one…and one that I spent quite a bit of time considering in the process of developing this article. Here’s what I came up with.

    Robots undoubtedly threaten human labor. They can work without rest, they are not prone to injury, they don’t go on strike, etc. This ultimately means that they are more efficient, and in many cases, less expensive. However, almost all robotic methods of fabrication are only usable offsite. They are not yet sophisticated enough to handle the complicated working conditions of a construction site. But considering the current pace of development, it is safe to assume that they will be in the foreseeable future (it is unclear whether that means 5 years or 20).

    The question, then, is when/if that happens, will robotic construction methods replace all human ones? I don’t think so. Surely, there will be certain types of for which this might be true. Skyscrapers, for example, could benefit greatly from a completely automated construction process as it would reduce time, cost, and danger. On the other hand, I can’t see robots replacing the local contractor (and his construction crew). So much architecture already exists in the world that is imprecise, definitely isn’t 3d modeled, and constantly evolving through its users. In this type of setting, not only is the precision of a robot entirely wasted, its requirement of a predetermined program could be very limiting.

    Yet both of these examples concern qualities brought up by robots in the first place. That assumes that construction, or making architecture, subscribes only to the capitalist model of faster and cheaper. In many circumstances, this is not the case. In an earlier draft of the editorial, I made this point with the Sagrada Familia, a building which gets it character from time and from making. It would not be the same building if robots built it in two years. This led me to a question: are there certain types of architecture that could not be built by robots? What about sacred architecture? And another: are there certain materials that require the human hand?Think of Ando’s concrete, Zumthor’s burnt stone.

    In fact, the argument in this article came out of my exploration of this question. How can we balance our limitless drive for technology with the unchanging fragility of humanity’s survival? By pointing our technological pursuits in the direction of humanity’s survival. We should use robotic architecture for those types of architecture that prioritize what its good at: speed, efficiency, and precision. We should not, however, try to employ robots comprehensively simply because they will save money. Like I said in the article, robots should not inspire architecture for the sake of robots, they should simply assist in architecture for the sake of people.


    • tc

      It’s inevitable that manufacturing and fabrication jobs will ultimately be replaced by automated robotic methods. This will result in a loss of manufacturing jobs, which is OK, because that is not typically where we, the US, are competitive. Where we will be competitive is in deigning the robotic systems and fabricating those new systems, all of which will also required human supervision. As we see technology overtake certain manual labor jobs, we’ll see a new market emerge for the design and construction of those systems.

      It’s already interesting to see, even in basic examples, on the job site. My current project had it’s floor slabs (of a 12 story tower) poured by a remote controlled robot. The operator just stood within 10 feet and drove the robot around spreading concrete, thus eliminating the additional 1-2 people that were previously needed and alot of back injuries that typically happen (and drive up the cost of construction) when holding the concrete hose. Human labor was still needed to spread the concrete evenly.

      agreed with your statements on the role of robotics in architecture. Like software, robots don’t make architecture. Ideas make Architecture. certain specific demands, like emergency housing could be tailored to the fabrication technique of employing robots, but I would imagine designing a museum around robot capabilities is letting the tail wag the dog.