“Complexity becomes an even bigger opportunity for architecture”

Mar 2, 2021

“Digital Building Technologies” sounds to many people familiar with the construction industry like an oxymoron. Benjamin Dillenburger, Professor for Digital Building Technologies at ETH Zurich, is here to change that. His research focuses on the development of building technologies based on the close interplay of computational design methods, digital fabrication and new materials. He searches for ways to exploit the potential of additive manufacturing for building construction – something we are similarly excited about over at Mighty Buildings. If you want to know why complexity is an opportunity for architecture, read on!

Today’s construction industry is not the first thing that comes to people’s minds when thinking of “digital”. Could you briefly explain your field of research?

My field of research can roughly be classified into two themes. On the one hand, I’m working on computational design methods that leverage artificial intelligence, machine learning, and evolutionary strategies to automatically generate and evaluate building designs.

On the other hand, I’m looking for ways to exploit digital fabrication technologies (eg. 3D printing, milling machines) and new materials (eg. polymers) for building construction. 

It’s important to understand that both research streams cannot be viewed in isolation. Making sure that computational design methods and digital fabrication technologies are aligned and harmonized is an integral part of my research focus.

How will your research change the construction industry?

Our research will ultimately increase the level of automation in construction and by that eventually increase the overall efficiency of the industry. 

Let’s use the DFAB house as an example: The idea was to erect a building relying almost entirely on robots and 3D printers. As part of the project my team was responsible for the design and fabrication of the slab. Instead of directly 3D printing the whole element, we decided to only print the formwork allowing us to use the geometric freedom and precision of 3D printing and pair it with the structural capacity of cast concrete. With this fabrication system, we could optimize the slab structurally by using computational design methods eventually reducing the amount of materials needed by 60%.

Beyond efficiency, our research will also allow architects to turn complexity into opportunity. Designs will be less constrained by limited building technologies and materials.

Let’s double-click on how computational design and digital fabrication will change the look of buildings. Are buildings of the same type (eg. multi-family homes) going to look similar, as the level of automation and optimization in construction progresses?

I think this scenario is not very likely for multiple reasons. In the construction industry the lot size is usually very small, in most cases only one. The boundary conditions vary from construction site to construction site: soil properties, orientation of the building, ambient environmental conditions, building codes to only name a few. Building designs need to be customized along those constraints, and will hence not be similar. 

We can apply analogies from other industries as well: in the automotive industry we see an incredibly high degree of automation and standardization leading to a high productivity in fabrication – still, not all cars look similar. 

In the context of construction, “standardization” has a negative connotation for many people – maybe because they associate the term with “Plattenbauten” (click here for translation) and everything that comes with it. 

As a matter of fact we are already using standardized parts beneath the surface of a building for good reasons. However, to continuously increase the customer acceptance of standardization and automation in construction we need to ensure the flexibility for customization. Digital technologies can allow for an efficient systematization, without the need to rely on fixed modules. Standards can become more fluid and open. 

We can expand the use of standardized components and processes without compromising on flexibility of customization.

So we are not going to see fully standardized blueprints for buildings like hospitals, schools, etc.?

I expect this to happen only to a certain degree and for some building types. I believe we will see an increase in systematization of construction, but those systems need to stay flexible. Again, optimizing a building design is essentially a multi-parameter optimization with project-specific constraints. In an ideal case we have full transparency on how a specific building design affects the manufacturability, CO2 footprint and total cost of ownership of a building. Based on that transparency we can then perfectly fit a building design with the lowest-possible cost, time and CO2, instead of applying a “one-size fits all” solution that actually wouldn’t fit any project.

Speaking of building design: how will the job profile of architects change given that many tasks may be automated?

That is a far-reaching question which will have multiple answers depending on whom you ask. 

What’s safe to say is that the understanding of all trades involved in a construction process and their respective intersections will be increasingly important for architects. 

Tech-savvy architects will leverage digital technologies to enhance their focus on design work by automating repetitive legwork. In a nutshell: when being done right, complexity is an opportunity for architects.

Benjamin, thanks a lot for your time! What’s the easiest way to learn more about your work and your projects?

Take a look at our website.

You might also like

“Orchestrated supply chains are the key to unlock efficiency in the AEC industry”

Jan 26, 2021

There aren’t many professorships specifically dedicated to innovative and industrial construction. Daniel Hall at ETH Zurich holds one of them. The overarching theme of his research is to enhance governance, productivity, and innovation in construction projects. His fascination for innovation in construction supply chains started when doing his PhD at Stanford and makes him a perfect expert for us to talk to. That’s what we did. If you want to learn more about his research, construction’s first principles and how to design and construct a house in seven days, you may want to read on.

Innovative and Industrial Construction is not a professorship that can be found at many universities. Why is it needed?

The architecture, engineering and construction (AEC) industry is characterized by an extreme, three-dimensional fragmentation of the stakeholders involved. It’s not only the vertical fragmentation between the different stages of a project, design and construction for example. In each of those stages you can also observe a horizontal fragmentation through the different trades contributing to a project. What drives complexity even further is the longitudinal fragmentation: apart from the fact that the location of a construction site changes project by project, teams disband after the job is finished, making it difficult to capture best-practices and leading to a “learning disability”.

That perfectly explains the state of our industry …

Exactly. The AEC industry is being held captive by the characteristics of its own structure. While other industries such as manufacturing were able to drastically increase their productivity, construction has not been able to keep up. Even worse, in construction the gross value added per hour worked dropped by 23% between 1990 and today.

A gigantic value creation opportunity over the next decades. How does your research help solve the AEC industry’s issues?

The overarching theme of my research is to enhance governance, productivity, and innovation in construction projects through a transformation from fragmented project delivery methods to new organizational models that integrate the supply chain of the AEC industry.

More specifically this includes research on new business models and circularity for industrialised construction, design for digital fabrication, and more recently also potential applications of the blockchain technology in construction.

Could you share a practical example of your research?

Our new project called “7 Day House” is a great example. We are asking the question, “What would it take to design and construct a house in seven days?” It’s important that we’re saying “design” and “construct” because right now it is possible to design a house very quickly. It’s also possible to construct a house very quickly. But: having a customized design and a connected supply chain to fabricate the house in seven days is a great conceptual and theoretical challenge.

Where’s the breaking point?

That remains to be seen. What I can say is that we give ourselves one day for the design. To achieve that we are going to leverage augmented and generative design tools. The nitty-gritty part will then be to also digitally model the fabrication. The idea is to take rules and constraints from later stages of the value chain (eg. pre-fabrication, assembly and logistics) into consideration to avoid creating a design that is inefficient to build or not buildable at all.

Being one of the founders of the Stanford Industrialized Construction Forum you get a lot of exposure to the construction tech world. Are there developments you are specifically excited about?

In general, we see a lot of traction. When we did the first forum in 2014 we were a small bunch of 30 people. That has changed. Like the venture capital and large financial investments that flow into that space, too. 

We’ve recently seen many prefab ventures that optimize and orchestrate their supply chains by vertically integrating. This model has its challenges with asset utilization being one of them.
What we are now seeing is that startups are starting to orchestrate construction supply chains not by vertically integrating but by looking at it from an “Digital Systems Integration” perspective. More precisely, these ventures develop an asset-light industry 4.0 system to orchestrate a deep network of suppliers.

Daniel, thanks a lot for your time! What’s the easiest way to learn more about your work and your projects?

Join the Industrialized Construction Forum on January 26th and 27th!

You might also like

“The belief that every building is unique has been an excuse not to industrialize construction”

Klaus Freiberg is an experienced blue chip CxO with a deep treasure trove of operator’s knowledge in the real estate sector. As COO of Vonovia, Europe’s largest residential real estate company, he insourced and operated the largest amount of skilled construction workers in the country – and reaping lean construction efficiencies from it. His heart beats to make companies customer-focused in real estate – and optimize the supply chains to get there. That is why we are happy to have him on board and talked to him about excuses, first principles, and the potential of standardization in construction.

You’ve worked in the built environment for quite some time. Looking at the industry in 2020, what is your state of mind?

We certainly are in a challenging situation. Construction is one of the biggest polluters globally, and the annual productivity growth has remained flat since 1945. And while I think everyone has recognized and accepted the challenges, not much has happened yet.

We have clear, unambiguous energy and CO2 reduction requirements from politics and society. But we are on no trajectory to meet them. Something needs to happen. That is also true concerning costs. And it is increasingly difficult to find qualified workers that build today’s buildings.

What are the underlying problems?

Today’s industry just isn’t industrialized or standardized at all. There is a reason that construction’s productivity dropped by 27 percent in the last twenty years, while agriculture or manufacturing went the other way. Honestly, when I started working in this sector some twenty years ago, I thought we would be much further in this regard as we are today. In the end, construction is still a ‘crafted’ process, not an industrialized one. That’s a pity.

But doesn’t construction have quite unique first principles that make it much harder to do so?

Of course. You will always have a unique location and assembly on-site, which means that your environment is much harder to control as a car factory. But the belief that every building is unique has been an excuse not to industrialize construction for too long. It’s just not true.


…depending on the origin, function, and date a building was constructed, you can see quite a lot of similarities and the same underlying logic. Just take a two or three-room flat built in 1970 Germany. You won’t find any round windows. They are 60 x 80, 80 x 100, or 100 x 120 (centimeters), and you have a balcony door, which is not oval. That’s for sure. The size of bathrooms only differs by 12 percent and is around 8 square meters. I bet you. Or take non-residential buildings. For decades, we have discussed the bad shape of our school buildings, especially sanitary rooms. But I don’t see why we can’t set up a standardized renovation approach for it. They all have pretty much the same size. So why don’t we pre-fabricate the sanitary rooms at low costs and make the final assembly on-site?

All I want to say: We should start focusing on the similarities of buildings instead of where they differ. The potential of standardization is tremendous. And it’s an answer to the shortage of skilled workers, too…

…which is another key challenge construction faces.

Absolutely. The young just don’t want to get their hands dirty. Only three percent of the 18-25 year-olds are interested in joining the construction trades. That is why we already are thousands of workers short. It’s a tough job, but we need to change the perception of the construction industry to attract young and motivated people.

That certainly is a tough job, too. What else can be done?

Well, we need to increase the productivity of a single worker. One way is augmented reality. If we have fewer experts in any given trade, we need to make the best use of the experts we have. If a worker is unable to fix a heating from 1982, why can’t he get support from an expert who is connected over data glasses? Another answer can be exoskeletons. I must admit I had my doubts, but recent solutions seem to make it right for lightweight and the high power to weight ratio while still being affordable. Human-machine collaboration is clearly something we should look into.

But as I said before, the most promising way to increase productivity is standardization.

Could you give an example of how you think standardization improves productivity?

Sure, let’s take the bathroom again. Usually, a craftsman comes with a toilet, a sink, 20 meters of copper pipe, and some bits and pieces to set it up in a unique way, which takes time. Why don’t we offer the whole module pre-fabricated so that two workers can lift and easily plug in?

If you look at how hotels are built these days, you get the idea. No one deals with interior construction on site anymore. Once the skeleton is complete, they start to slide in the nearly complete rooms, including interior walls and bathrooms — highly efficient and low costs. 

Don’t get me wrong. I am not saying these ideas are brand new, but we finally have to get going.

Sounds good, but how do we get there?

Well, if you look at the industry, you see some resilience to change. While in other sectors, half of all companies have been acquired or gone bankrupt in the last 20 years, 85% of construction companies are still there. Unchanged and undisrupted. That shows us that the sector won’t change from within. We need outside innovation to lift these potentials and apply the right tech. And we need venture capital to accelerate that change.

You might also like

Text Message Interview with Mighty Buildings

In our text message interview, Sam Ruben, Co-Founder and Chief Sustainability Officer at Mighty Buildings, explains how they are solving the housing crisis with houses that are entirely 3D-printed.

Mighty Buildings strives to make homes more affordable by automating up to 80% of the entire construction process. Compared to an average house in California, their homes cost as much as 45% less. With their innovative material they can print a 350 square foot house in just 24 hours, which also answers construction’s labor shortage. Building the “Mighty way” also has environmental advantages. They use a single material, waste is prevented as channels for plumbing and electrical are printed and are able to produce energy-efficient forms of a house. Lower costs, scalable material and environmental benefits. That’s what we love about Mighty Buildings. 🏡❤️

You might also like

“The whole sector is deeply undermanaged”

Stefan Kirsten is an industrial veteran with a background as a German blue chip CFO at companies like Metro Group, Thyssenkrupp as well as Vonovia, the largest German residential real estate company. He currently serves as a non-executive director in both growing start-ups and mature companies. Besides teaching as a professor, he just joined the Foundamental advisory board. That’s why we called him and asked for his take on what’s going in the building world. Here are his answers on undermanaged industries, rising costs and the labor shortage in construction. Definitely worth a read (if you ask us).

Stefan, you’ve worked in the building world on multiple continents. Looking at these industries in 2020, what is your state of mind?

These are challenging times. For all of us, and the building world makes no difference. If we look at the planet and the decades ahead, we can say that demand will be there. It’s simply demographics. It will obviously differ from continent to continent, but we are about to add 13’000 new homes every day for the next 40 years. That’s net new – every single day…

The main problem, however, is that the sector is deeply ‘undermanaged’. We need to address that with all the radical means necessary – especially if you look at bringing down carbon emissions.

When you say the sector is undermanaged you mean…

It basically means that you have a high degree of fragmentation. Just look at the numbers: 80% of value-add in construction comes from companies with less than 250 employees. And it’s no surprise if you look at what makes construction so unique. Construction sites will always be geographically dispersed. That’s for sure. Ten construction sites will always be in ten locations. With all the characteristics and regulations prone to the area of the site. Also, the assembly itself will always be distinctly local. Fully-assembled structures are just too heavy to be transported. And you can’t transport cement over 4000 kilometers. It just doesn’t make sense. Compare that to the car industry for example: they have controlled environments where the car is assembled mostly by machines as they can ship it easily wherever they want. That’s why they have seen a steep increase in automation, tech-adoption and, hence, productivity. And construction did not.

Which leads to the spiraling costs seen in past decades…

Exactly. Low productivity is one reason that drives costs. As well as local constraints on site and regulations that make it nearly impossible to build in developed countries (all implemented for the right reasons, but also driving costs in the end). But what shouldn’t be overlooked when talking about costs is the looming construction labor crisis. 

In many markets we have an aging labor force. Up to 40% of the US construction workforce today will be retired in ten years. The problem is that the young aren’t keen on joining the construction trades. Meaning that the US alone will be short of a couple of millions of workers by 2030. And it’s pretty much the same in all other mature markets.

Doesn’t that mean we will have to build the most complex projects we’ve ever built with the least experienced workforce we’ve ever had? What can we do about it?

That certainly is one way of putting it. Addressing it won’t be easy, but we have to. We have to better utilize the workforce. That means we will have a higher disparity between skilled and unskilled workers. Let me give you an example: when you refurbish a flat, the refurbishing crew should not carry the materials into the flat. That’s what a freight forwarder does. Because these guys can do bathrooms, and they can only do bathrooms, but they do them in 72 hours. And they do them very well. By letting the experts do their job, and only their jobs, productivity goes up. 

Another way of increasing output of a single worker is standardization. That combats complexity and is one way to not only increase productivity of the unskilled, but also make life easier for everyone involved. And, of course, there is the point of best managing all the different parts of a construction supply chain and optimizing processes. It sounds so simple, but there is so much room for improvement, believe me.

We do. But how to change an industry with such a high fragmentation?

Good question. I mean the industry isn’t solving it from within. Too fragmented and too little spend on R&D. We need to attract outside innovators, being it tech founders or industry experts that are fed up with the status quo. Let’s make use of the technology that is certainly out there and use it in every step of the supply chain. Once done, you can much better manage the industry. 

But we must also get better in learning from the past. I don’t want to drop any names, but even big firms start every project as if they are reinventing it. And each time when you hear the word experience, it simply means knowledge which is in brains, but not in systems, structures and processes. And this is something where this industry is significantly underdeveloped. I mean if you have built, for instance, the Olympic Games in Sochi and you want to build them in Tokyo, I doubt that they will take more than 2% of the knowledge and experience from Sochi over to Tokyo. And that’s stupid because it’s still a stadium, isn’t it? 

Let’s change that with bright minds fueled by venture capital.

You might also like

“There is no other sector that holds such a big lever for climate impact”

Gaurav Sant is a professor for civil and environmental engineering as well as materials science and engineering at the University of California, Los Angeles (UCLA). But his interests go beyond the written word as he is also founder of CO2Concrete, a deep-science venture working on manufacturing a low-carbon concrete-equivalent material that releases 75% less carbon than conventional techniques and can store waste carbon dioxide (CO2) too. We talked to him about democratizing data in construction, the future of concrete and whether the building world can deliver on the Paris Agreement.

Gaurav, construction and mining are often overlooked but account for 17% of global CO2 emissions. Meaning that there is obviously a lever here. What is the most promising way to bring down these emissions?

Good question. Just look at the construction phase of a building: over 90 percent of emissions come from the building materials that are used in the building’s construction. That means we have to do something about the production of these materials, innovate and curate low-carbon substitutes, and significantly reduce waste.

Both construction and mining produce materials that are used by other downstream sectors that are effectively hostage to the upstream materials that are available. That means we have to consider the entire supply and value chain, from better decisions in pre-construction to sustainable or circular materials. For all of that, technology plays a key role. And I still believe technology is our best bet to reduce the cost of, and enable decarbonization. But technology won’t get us anywhere if we aren’t able to cultivate and build a social conscience, where we as society demand low-carbon materials and low-carbon products. Here, all of us as individuals are responsible for fostering much-needed change.

Besides building up social awareness and given the fact that over 90 percent of emissions in construction come from building materials: where do we stand with carbon-neutral substitutes?

I think we’ve got a lot of promising substitutes starting to come about. We’ve been building with “soil and straw” for a long time. However, these options are unscalable, as we’ve demonstrated over and over again. For materials like steel or cement, carbon-neutral substitutes already exist in research labs and on a smaller-scale. The question is, how do we really start to scale these solutions and how do we make them accessible economically? How do we really raise the money that is required to bring these solutions to market? How do we make a lot of big bets with an acceptance and a willingness that not everything will work? We’re going to need to deploy a lot of money to find out. Extremely hard things to think about, because, again, this is an extremely reluctant sector that does not prioritize investments in ‘true innovation’. Don’t forget: This is an industry that promotes tradition rather than transformation. But we have to make big bets and accept failure. It’s the only pathway to success, especially with something as radical as carbon-neutral materials.

The funny thing is everyone talks about becoming carbon neutral by 2050. This is 30 years – which is actually a really short period of time. But the money simply isn’t’ flowing in. Why? Because, as a good friend presciently pointed out to me a few years ago, in these commodity sectors, “everyone wants to be the first person to be second!”

Well, then let’s also talk about becoming carbon neutral in 2050. What does that mean for a material like concrete, will we still build with it in 30 years time?

I think the answer is unquestionably, yes. It might be a different composition, color and form of concrete than we use today, which is based on OPC aka ordinary portland cement. But concrete is going to be a part of our future, as it has been in the past and present. The reasons are simple: low-cost, large abundance, and our experience in manufacturing these materials on a large scale around the world. 

But, we have to be pragmatic. OPC-based concrete has been our material of choice for construction for nearly 150 years. A transition is not going to happen overnight. But, for this disruptive transition to occur we need both incremental and transformative innovation. They will work hand in hand over the next few decades to displace and reimagine concrete as we know it. This will leverage a suite of solutions including AI-based materials design and optimization, carbon abatement solutions, and turning CO2 emissions into construction materials such as CO2Concrete.

All that said, we are going to see enormous changes in concrete as a material, how we specify it, and how we use it. Accepting that change and accepting that there’s going to be a new and different concrete to look forward to is a big part of the ‘reset’ that we need to build into our industrial and construction base.

Looking at the need to make construction more efficient and even apply AI, does the sector have the infrastructure ready to do that?

Not really. At least not in a sufficiently democratized manner. It’s a question of infrastructure, data and processes aka algorithms. In some cases data exists, in many others it doesn’t. By solving problems in construction with digital technologies, new data pools emerge everywhere. That is great. But these data pools need to be organized. That is key for me. There needs to be a willingness to share and connect these data pools, because that is when real value is created.

I would argue that really developing and curating the data and being able to develop approaches to democratize the power of data is really where the opportunity lies. However, the construction industry is very reluctant to share data either for reasons that are based on  perceived competitiveness or that are just purely misunderstood.

With regard to AI, we have seen tremendous progress, but much more so in the ‘soft space’ of managing and building new efficiencies and workflows. We have not really looked at using AI to drive actual interventions in physical manufacturing processes. I think this is really where the other big leverage really lies. Just imagine if you can produce every building in the world using 20 percent less of every material that’s used. That’s a big breakthrough and purely leveraged by accessing industrial efficiencies. This works towards the premise that the easiest, and cheapest, carbon abatement that we can make happen is the carbon we don’t emit to begin with.

Isn’t that especially true as the most of construction is yet to come? Studies predict that we are going to add the floor area of Japan to the planet each year. Where does that leave your optimism? Do you think that construction and mining can deliver on the Paris Agreement despite this sheer endless demand for buildings?

Yes, 100 percent. There is no doubt that technologically we have the innovation capacity that’s needed to make this a reality. The question is, do we have the willingness financially to make this come to life and scale it to industrial levels (i.e. billions of tonnes of material production per year)? To me, carbon is not a technological challenge, it’s an economic one. And we will only find an effective answer by scale-up, scale-out, and deep investments in cross-sectoral solutions. We should not so much ask what is the lowest cost for carbon abatement in specific sectors, but rather on a per capita basis. That is also key to me because carbon dioxide is not a sector specific issue; it’s a societal one!

Given the need to translate technologies and research to industrial scale rather soon, what is your recipe for success here?

My recipe is bold and bright minds, venture capital, supportive and timely regulation and a shift in the corporate world. Let’s unpack that. First, we need to draw attention to industries that are the big polluters and convince founders and soon-to-be founders to move into these spaces; in spite of the perception that these sectors are entrenched, and hence “hard to disrupt”. Second, there’s a huge role for venture capital to back new ventures, make big bets, and make sure there is product-market-fit on an industrial level. Ensuring the industrial technology readiness level, so to speak. We have recently seen an inflection of venture capital towards construction and mining, but it’s not even a drop in a bucket. Which is strange, as there is no other sector that is so big, undisrupted and holds such a big lever for climate impact; and has the potential to deliver double-/triple-bottom line impact like few sectors could.Third, we need strong forward-looking, and decisive regulation and policy making, and we need an absolute modification and transformation in corporations around the world and how they think. It’s nice to establish an aspirational pledge regarding how much Corporation X wants to reduce its CO2 emissions. But, often these pledges are just that: aspiration. On the other hand, if corporations (and their shareholders) linked reductions in CO2 emissions to executive compensation, we’ll see that decarbonation is going to be authentic, and it will happen much, much faster. So it’s really about aligning incentives across executive management, consumers (of a corporation’s products or services), and society as a whole.

Thanks Gaurav for taking the time.

Note: Gaurav Sant is a member of the Foundamental advisory board

You might also like